Literature Survey
------Lateral Dynamics of Commercial Vehicle Combinations-A Literature Survey(Vehicle System Dynami
As compared with single vehicle units, the road vehicle trains are characterized by some specific features among which particularly the following should be noted (Fig. 2): a ) jackknifing b) trailer swing c ) trailer lateral oscillations Under conditions of high vehicle velocities the articulated vehicles tend to yaw oscillations causing large amplitudes of lateral trailer oscillations. The combinations truck-trailer are more prone to lateral oscillations than the tractorsemitrailer sets. Under braking conditions on wet slippery roads the blocking of one or more wheels causes jackknifing or a swing of the semitrailer or trailer. The combination TS (tractor-semitrailer) appears to be more sensitive to these effects. The tractor has to transmit a part of the mass of the semitrailer and, consequently, also a part of the braking power indispensable for braking the semitrailer, and thus considerable forces are exerted at the kingpin connecting the tractor and the trailer. When the tractor rear wheels are blocked, the tractor swings around the kingpin. In case the semitrailer wheels get blocked, the semitrailer starts to yaw by skidding around the axis of the connecting mechanism (i.e. the fifth wheel). The lateral stability and braking performance of articulated vehicles are problematic, in particular for the following reason: owing to forces in the connection
Active Learning Literature Survey
Active Learning Literature SurveyAnita KrishnakumarUniversity of California,Santa CruzDepartment of Computer Scienceanita@June05,2007AbstractThe most time consuming and expensive task in machine learning is the gatheringof labeled data to train the model or to estimate its parameters.In the real-worldscenario,the availability of labeled data is scarce and we have limited resources tolabel the abundantly available unlabeled data.Hence it makes sense to pick onlythe most informative instances from the unlabeled data and request an expert to pro-vide the label for that instance.Active learning algorithms aim at minimizing theamount of labeled data required to achieve the goal of the machine learning task inhand by strategically selecting the data instance to be labeled by the expert.A lot ofresearch has been conducted in this area over the past two decades leading to greatimprovements in performance of several existing machine learning algorithms andhas also been applied to diversefields like text classification,information retrieval,computer vision and bioinformatics,to name a few.This survey aims at providingan insight into the research in this area and categorizes the diverse algorithms pro-posed based on main characteristics.We also provides a desk where different activelearning algorithms can be compared by evaluation on benchmark datasets.1IntroductionThe central goal of machine learning is to develop systems that can learn from experience or data and improve their performance at some task.In many natural learning tasks,this experience or data is gained interactively,by taking actions,making queries,or doing experiments.Most machine learning research,however,treats the learner as a passiverecipient of the data to be processed.This passive approach ignores the learner’s ability to interact with the environment and gather data.Active learning is the study of how to use this ability effectively.Active learning algorithms have been developed for classification, regression and function optimization and is found to improve the predictive accuracy of several algorithms compared to passive learning.2Major ApproachesThe three major approaches to active learning algorithms are as follows.•Pool-based active learning:As introduced in Lewis and Gale(1994),the learner is provided with a pool of independent and identically distributed unlabeled instances.The active learner at each step chooses an unlabeled instance to request the label from the expert by means of a querying function.This is also called as selective sampling.•Stream-based active learning:The active learner,for example in Freund1997,is presented with a stream of unlabeled instances,from which the learner picks an instance for labeling by the expert.This can be visualized as online pool-based active learning.•Active learning with membership queries:Here,as described in Angluin1988,the active learner asks the expert to classify cases generated by the learning systems.The learner imposes values on the attributes for an instance and observes the re-sponse.This gives the learner theflexibility of framing the data instance that will be the most informative to the learner at that moment.3Characteristics of Active Learning AlgorithmsWe have taken into account several key features that have been addressed in many of the proposed active learning algorithms to compare the effect of each characteristic on the overall performance of the algorithm.3.1Ranker considerationActive learning algorithms may or may not depend on a ranker function to pick the train-ing instance for expert labeling.Several algorithms proposed use support vector machines (SVM),logistic regression,naive Bayes,neural networks,etc.In this section we look atactive learning algorithms from the perspective of the ranker function used in the data instance selection process.Lewis and Gale(1994)describe an uncertainty sampling method where the active learner selects instances whose class membership is most unclear to the learner.Different defini-tions of uncertainty have been used,for example the Query-by-Committee algorithm by Seung et al.(1992),picks those examples for which the selected classifiers disagree,to be labeled by the expert.The authors suggest that their algorithm can be used with any classifier that predicts a class as well as provides a probability estimate of the prediction certainty.Cohn et al.(1995)describe how optimal data selection techniques can be applied to statistically-based learning algorithms like a mixture of Gaussians and locally weighted regression.The algorithm selects instances that if labeled and added to the training set, minimizes the expected error on future test data.The authors show that the statistical models perform more efficiently and accurately than the feedforward neural networks. Similar querying functions have been proposed by Tong and Koller(2000),Campbell et al.(2000)and Schohn and Cohn(2000),called Simple which uses SVMs as the induction component.Here,the querying function is based on the classifier.The algorithm tries to pick instances which are the most informative to the SVM-the support vectors of the dividing hyperplane.This can be thought of as uncertainty sampling where the algorithm selects those instances about which it is most uncertain.In the case of SVMs,the classifier is most uncertain about the examples that are lying close to the margin of the dividing hyperplane.Variations of the Simple algorithm-MaxMin and Ratio methods have been proposed by Tong and Koller(2000),which also use SVM as the learner.Iyengar et al.(2000)present an active learning algorithm that uses adaptive resampling (ALAR)to select instances for expert labeling.In the work described,a probabilistic classifier is usedfirst to determine the degree of uncertainty and then decision trees is used for classification.The experiments considered use the ensemble of classification models generated in each phase or a nearest neighbor(3-nn)as the probabilistic classifier. Roy and McCallum(2001)describe a method to directly maximize the expected error rate reduction,by estimating the future error rate by a loss function.The loss functions help the learner to select those instances that maximize the confidence of the learner about the unlabeled data.Rather than estimating the expected error on the full distribution,this algorithm estimates it over a sample in the pool.The authors base their class probability estimates and classification on naive Bayes,however SVMs or other models with complex parameter space are also recommended.Baram et al.(2003)provide an implementation of this method on SVMs,andfind it to be better than the original naive Bayes algorithm. Logistic regression is used to estimate the class probabilities in the SVM based algorithm. Baram et al.(2003)propose a simple heuristic based on“farthest-first”travel sequences for active learning called Kernel Farthest-First(KFF).Here the active learner selects thatinstance which is farthest away from the current labeled set and can be applied with any classifier learning algorithm.The authors present an application of KFF with an SVM. Mitra et al.(2004)present a probabilistic active learning strategy for support vector ma-chine learning.Identifying and labeling all the true support vectors guarantees low future error.The algorithm uses the k nearest neighbor algorithm to assign a confidence factor c to all instances lying within the boundary,close to the actual support vectors and1−c to interior points which are far from the support vectors.The instances are then chosen probabilistically based on the confidence factor.Nguyen and Smeulders(2004)offer a framework to incorporate clustering into active learning.A classifier is developed based on the cluster representatives and a local noise model propagates the classification decision to the other instances.The model assumes that given the cluster label,the class label of data instance can be inferred.The logistic regression discriminative model is used to estimate the class probability and an isotropic Gaussian model is used to estimate the noise distribution,to propagate information of label from the representatives to the remaining data.A coarse-to-fine strategy is used to adjust the balance between the advantage of large clusters and the accuracy of the data representation.Osugi et al.(2005)propose an active learning algorithm that balances the exploration and exploitation while selecting a new instance for labeling by the expert at each step. The algorithm randomly chooses between exploration and exploitation at each round and receives feedback on the effectiveness of the exploration step,based on the performance of the classifier trained on the explored instance.The active learner updates the probability of exploring in the next phases based on the feedback.The algorithm chooses between the active learners KFF(which explores)and Simple(which exploits),using SVM light as the classifier,with a probability p.If the exploration is a success,resulting in a change in the current hypothesis,then p is maintained with a high value,encouraging more exploration, else p is updated to reduce the probability of exploration.3.2Computational complexityComputational cost is an important factor to be considered in an algorithm.If an algorithm is computationally very expensive,the algorithm might be infeasible for certain real-world applications which are time sensitive.In this section,we consider the time complexities of the above discussed active learning algorithms infinding an optimal instance for labeling. The active learning algorithms with mixture of Gaussians and locally weighted regres-sion proposed by Cohn et al.(1995)performs more effectively than the feedforward neural networks where computing variance estimate and re-training is computationally very expensive.With the mixture of Gaussians,training depends linearly on the number of data instances,but prediction time is independent.On the other hand,for a memory-based model like locally weighted regression,there is no training time,but prediction costs exist.However,both can be enhanced by optimized parallel implementations.The authors Tong and Koller(2000)suggest that the Simple margin active learning al-gorithm is computationally quite efficient.However,improvement gains can be obtained by querying multiple instances at a time as suggested in Lewis and Gale(1994).But the MaxMin and Ratio methods are computationally very expensive.Active learning with adaptive resampling(Iyengar et al.,2000)is computationally very expensive because of the decision trees used in the classification phase.The number of phases,number of points chosen to be labeled and number of adaptive resampling rounds also add to the computational cost,hence the authors have chosen these parameters based on computational complexity and accuracy considerations.The computational complexity to implement the algorithm proposed in Roy and McCal-lum(2001)is described as“hopelessly inefficient”.However,various heuristics approxi-mations and optimizations have been suggested.Some of these approximations are gen-eral and some for the specific implementation by the authors on naive Bayes.The computational complexity of Kernel Farthest-First algorithm of Baram et al.(2003) is quite similar to the Simple margin algorithm.Simple computes the dot product for every unlabeled instance which takes the same time as computing the argmax for KFF.The probabilistic active support vector learning algorithm proposed by Mitra et al.(2004) is computationally more efficient than even the Simple margin algorithm by Tong and Koller(2000),as presented in the comparison by the authors.The active learning using pre-clustering algorithm proposed by Nguyen and Smeulders (2004),use the K-medoid algorithm for clustering as it captures data representation bet-ter.But the K-medoid algorithm is computationally very expensive when the number of clusters or data points is large.However,some simplifications have been presented to reduce the computational cost.The algorithm proposed by Osugi et al.(2005),uses the active learners,Simple or KFF. Hence the time complexity depends on those algorithms.Simple and KFF have similar time complexities and hence,this algorithm has the sample complexity of those algo-rithms per round.3.3DensityIn real-world applications,the data under consideration might have skewed class distri-butions.Some classes might have larger number of samples,and hence a higher density than the other classes.Some classes might have very few instances in the dataset andhence a low density.In this section we analyze whether the active learning algorithms in discussion consider the density of the classes in the dataset,while selecting instances for labeling.The statistical models of Cohn et al.(1995)selects that instance that minimizes the ex-pectation of the learner’s variance on future test set.The instance selection method is independent of density considerations.The Simple algorithm of Tong and Koller(2000),Campbell et al.(2000)and Schohn and Cohn(2000),picks those instances which are close to the dividing hyperplane.Density of the class distribution is ignored here.The ALAR algorithm of Iyengar et al.(2000)also selects instances for expert labeling ignoring the density of samples,as it only considers the degree of uncertainty of the classifier.The algorithm of Roy and McCallum(2001)queries for instances that provide maximal reassurance of the current model.Hence,it does not depend on the class density distribu-tion.The KFF algorithm of Baram et al.(2003)selects those instances that are the farthest from the current set of labeled instances,which does not really take into account the density of samples.The probabilistic active support vector learning algorithm by Mitra et al.(2004)does not take into account the density while querying for instances.The data selection criterion of the active learning algorithm with pre-clustering by Nguyen and Smeulders(2004),gives priority to samples which are cluster representatives,and chooses the ones belonging to high density clustersfibeling of high density clusters contribute to a substantial move of the classification boundary,and hence the algorithm clusters the data into large clusters initially.And once the classification boundary between the large clusters have been obtained,the parameters are adjusted to obtainfiner clustering for a more accurate classification boundary.The active learning algorithm proposed by Osugi et al.(2005)explores for new instances using KFF,which does not consider the density of the class distribution.The exploitation phase uses the Simple algorithm which again does not consider the density of samples.3.4DiversitySome active learning algorithms can have added advantage if they take into account the diverse nature of instances in the dataset.The classifier developed will perform well when trained with dataset that has different kinds of samples that represent the entireRanker Computational Density DiversitycomplexityAlgorithm1 x xAlgorithm2 x x xAlgorithm3 x xAlgorithm4 x xAlgorithm5x x xAlgorithm6 x xAlgorithm7Algorithm8x x xTable1:Summary of characteristics of the active learning algorithms in study distribution.The algorithms by Cohn et al.(1995),Tong and Koller(2000),Campbell et al.(2000)and Schohn and Cohn(2000)do not consider the diversity of the samples in the labeled set used for training the classifier.They just select the examples that optimize their criterion, which is minimizing the variance in Cohn et al.’s model and choosing the most unclear example to the classifier in the Simple algorithm.The ALAR algorithm by Iyengar et al.(2000)and the algorithm by Roy and McCallum (2001)also do not select instances based on their diversity.The KFF algorithm by Baram et al.(2003)select those instances that are the farthest from the given set of labeled examples.Intuitively,this picks the instance from the unla-beled which is most dissimilar to the current set of labeled examples used for training the classifier.The probabilistic algorithm by Mitra et al.(2004)selects samples that are far from the current boundary with a confidence factor c.This kind of helps the active learner to pick instances in the dataset that are diverse in nature.The active-learning algorithm by Nguyen and Smeulders(2004),selects diverse samples as it gives priority to samples which are cluster representatives,and each cluster represents a different group of data instances.The algorithm by Osugi et al.(2005),uses KFF in the exploration phase,which considers the diversity of the dataset while selecting the next instance for expert labeling.3.5Close to boundaryInstances lying close to the decision boundary generally contribute to the accuracy of the classifier,as in the case of support vector machines.Hence,those samples lying close to the boundary convey a lot of information regarding the underlying class distribution.In this section we see if the algorithms under study consider the instances lying close to the boundary for expert labeling.The statistical algorithms of Cohn et al.(1995)selects instances that minimize the vari-ance of the learner on the future dataset,and this might be equivalent to picking those instances close to the current decision boundary.The Simple algorithm described queries for instances that the learner is most uncertain about and this leads to choosing samples that are close to the classifier’s decision bound-ary.The ALAR algorithm using3-nn classifier for thefirst task of the determining the degree of uncertainty,minimizes the cumulative error by choosing the instances that are misclas-sified by the classifier algorithm in the second task,given the actual labels are those given by thefirst algorithm.This chooses the samples that are close to the current decision boundary.The active learning algorithm by Roy and McCallum(2001)tries to pick samples that provide maximum reassurance of the model and hence does not pick examples close to the boundary.The KFF algorithm also does not choose samples close to the boundary,it queries for the sample farthest from current labeled training set.The algorithm by Mitra et al.(2004)tries tofind the support vectors of dividing hyper-plane and hence considers the samples lying close to the decision boundary.The active learning with pre-clustering algorithm,tries to minimize the current error of the classifier.This leads to choosing those samples lying on the current classification boundary as they contribute the largest to the current error.The algorithm by Osugi et al.(2005)queries for the samples lying close to the boundary during the exploitation phase,using the Simple active learning algorithm.3.6Far from boundarySome active learning algorithms might query for instances that are far from the current decision boundary,as these examples can help to reassure the model as well as give a chance to explore new instances which might be very informative.The algorithm by Cohn et al.(1995),the Simple algorithm,the ALAR algorithm and theKFF do not query for samples lying far from the current decision boundary.The algorithm proposed by Roy and McCallum(2001)queries for examples that reduce the future generalization error probability.This leads to picking examples that reassure the current model and the samples lying far from the boundary are chosen by the algorithm as the learner is most sure about the labels for those samples.The active learning algorithm by Mitra et al.(2004)queries for samples lying far from the boundary using the confidence factor c,which varies adaptively with iteration.The algorithm proposed by Nguyen and Smeulders(2004)picks instances that lie close to the boundary for expert labeling,not the ones far away as they do not contribute towards the current error of the classifier.The algorithm by Osugi et al.(2005)queries for the samples lying far from the boundary during the exploration phase using the KFF algorithm.3.7Probabilistic or uncertainty of rankerHere we consider whether the ranker used in the active learning algorithm is probabilistic and whether the uncertainty of the ranker is used to query for new instances for expert labeling.Most algorithms studied here in this survey use a probabilistic ranker or uncer-tainty of the ranker to pick the samples for labeling.The active learning algorithms with statistical models by Cohn et al.(1995)uses prob-abilistic measures for determining variance estimates.The Simple active learning algo-rithm use uncertainty sampling to pick the instance that is most unclear to the learner. The ALAR algorithm that uses the ensemble of classifiers generated in the second task is probabilistic and chooses those samples that are misclassified by the learner.In the algo-rithm by Roy and McCallum(2001)uses probabilistic estimates using logistic regression to calculated the expected log-loss.The KFF algorithm does not depend on the probabilistic or uncertainty of the ranker. The algorithm of Mitra et al.(2004)uses the confidence factor c to pick the samples for labeling which is correlated with the selected samples in the labeled training set.The algorithm of Nguyen and Smeulders(2004)uses logistic regression with a proba-bilistic framework and also employs soft cluster membership to choose the sample for labeling.The algorithm by Osugi et al.(2005)considers probability measures for exploring based on feedback.3.8MyopicAn algorithm has a myopic approach when it greedily choose for instances that optimize the criterion at that instance(locally),instead of considering a globally-optimal solution. Most active learning algorithms choose a myopic approach as the learner thinks that the instance it selects for the expert to label,is the last instance that the expert is available for labeling.Most of the algorithms considered in this study adopt a myopic approach as they try to select the instance that optimizes the performance of the current classifier on the future test set.This is a major limitation in case of greedy margin based methods as the algo-rithm never explores if the examples lying far from the current decision boundary have more information to convey regarding the class distribution,which helps the classifier to become more effective.The statistical algorithm of Cohn et al.(1995)queries for the instance that minimizes the expected error of the model by minimizing its variance,which is actually myopic in approach.Similarly,the Simple algorithm by Tong and Koller(2000),Campbell et al. (2000)and Schohn and Cohn(2000)query for the instance that current classifier is most unclear about,at each iteration.Roy and McCallum(2001)also choose the instance that reassures the current model.The KFF algorithm by Baram et al.(2003)also chooses the example that is most different from the current dataset.However,some of these active learning algorithms select multiple instances to request label from the expert at each iteration instead of choosing just one instance.For example, the ALAR algorithm by Iyengar et al.(2000)and the probabilistic algorithm of Mitra et al.(2004)allows the learner to query the expert for labels of multiple samples at each instance.However this does not exactly globally-optimize the problem in hand.The algorithm of Nguyen and Smeulders(2004)gives priority to the cluster represen-tatives for labeling after an initial clustering.It also prioritizes examples from the high density clustersfirst for labeling.This gives the algorithm a kind of approach for global optimization by choosing diverse samples and high density cluster samples initially.But in the later stages the proposed method chooses those instances that contribute the largest to the current error.The algorithm by Osugi et al.(2005)also gives importance to exploring the dataset with a probability p,apart from just optimizing the current criterion.A high value of p encour-ages exploration and is maintained with a high value if the current hypothesis changes. Otherwise it is updated to reduce the probability of exploration at the next step.The value of p has an upper and lower bound,and hence there is always a chance of exploring and exploiting.Close to Far from Probabilistic/Myopicboundary boundary uncertaintyof rankerAlgorithm1 xAlgorithm2 xAlgorithm3 x not clearAlgorithm4xAlgorithm5x x xAlgorithm6 not clearAlgorithm7 x xAlgorithm8 xTable2:Summary of characteristics of the active learning algorithms in studyAlgorithm Authors Name1Cohn et al.(2000)Active learning with statistical models2Tong and Koller(2000)Simple MarginCampbell et al.(2000)Query learning with large margin classifiersSchohn and Cohn(2000)Less is More:Active learning with supportvector machines3Iyengar et al.(2000)Active learning with adaptive resampling4Roy and McCallum(2001)Active learning with Samplingestimation of error reduction5Baram et al.(2003)Kernel Farthest First6Mitra et al.(2004)Probabilistic active support vectorlearning algorithm7Nguyen and Smeulders(2004)Active learning with pre-clustering8Osugi et al.(2005)Balancing exploration and exploitationalgorithm for active learningTable3:Key to Table1&24ConclusionActive learning enables the application of machine learning methods to problems where it is difficult or expensive to acquire expert labels.Empirical results presented in the studied research papers indicate that active-learning based classifiers perform better than passive ones.In this paper we have presented a survey of several state-of-the-art active learning algorithms as well as the most popular ones.A detailed analysis of each algorithm has been made based on the characteristics of each algorithm,which gives an insight into the features each active learning algorithm considers while querying for instances for expert labeling.References[1]A NGLUIN,D.Queries and concept learning.Mach.Learn.2,4(1988),319–342.[2]B ARAM,Y.,E L-Y ANIV,R.,AND L UZ,K.Online choice of active learning algo-rithms,2003.[3]C AMPBELL,C.,C RISTIANINI,N.,AND S MOLA,A.Query learning with largemargin classifiers.In Proc.17th International Conf.on Machine Learning(2000), Morgan Kaufmann,San Francisco,CA,pp.111–118.[4]C OHN,D.A.,G HAHRAMANI,Z.,AND J ORDAN,M.I.Active learning withstatistical models.In Advances in Neural Information Processing Systems(1995),G.Tesauro,D.Touretzky,and T.Leen,Eds.,vol.7,The MIT Press,pp.705–712.[5]F REUND,Y.,S EUNG,H.S.,S HAMIR,E.,AND T ISHBY,N.Selective samplingusing the query by committee algorithm.Machine Learning28,2-3(1997),133–168.[6]I YENGAR,V.S.,A PTE,C.,AND Z HANG,T.Active learning using adaptive resam-pling.In KDD’00:Proceedings of the sixth ACM SIGKDD international conference on Knowledge discovery and data mining(New York,NY,USA,2000),ACM Press, pp.91–98.[7]L EWIS,D.D.,AND G ALE,W.A.A sequential algorithm for training text clas-sifiers.In SIGIR’94:Proceedings of the17th annual international ACM SIGIR conference on Research and development in information retrieval(New York,NY, USA,1994),Springer-Verlag New York,Inc.,pp.3–12.[8]M ITRA,P.,M URTHY,C.,AND P AL,S.K.A probabilistic active support vectorlearning algorithm.IEEE Transactions on Pattern Analysis and Machine Intelli-gence26,3(2004),413–418.[9]N GUYEN,H.T.,AND S MEULDERS,A.Active learning using pre-clustering.InICML’04:Proceedings of the twenty-first international conference on Machine learning(New York,NY,USA,2004),ACM Press,p.79.[10]O SUGI,T.,K UN,D.,AND S COTT,S.Balancing exploration and exploitation:Anew algorithm for active machine learning.In ICDM’05:Proceedings of the Fifth IEEE International Conference on Data Mining(Washington,DC,USA,2005), IEEE Computer Society,pp.330–337.[11]R OY,N.,AND M C C ALLUM,A.Toward optimal active learning through samplingestimation of error reduction.In Proc.18th International Conf.on Machine Learn-ing(2001),Morgan Kaufmann,San Francisco,CA,pp.441–448.[12]S CHOHN,G.,AND C OHN,D.Less is more:Active learning with support vectormachines.In Proc.17th International Conf.on Machine Learning(2000),Morgan Kaufmann,San Francisco,CA,pp.839–846.[13]S EUNG,H.S.,O PPER,M.,AND S OMPOLINSKY,H.Query by committee.InComputational Learning Theory(1992),pp.287–294.[14]T ONG,S.,AND K OLLER,D.Support vector machine active learning with applica-tions to text classification.In Proceedings of ICML-00,17th International Confer-ence on Machine Learning(Stanford,US,2000),ngley,Ed.,Morgan Kaufmann Publishers,San Francisco,US,pp.999–1006.。
Critique_the_Literature
CONCEPT 1. MAKING THE CASE FOR THE LITERATURE REVIEW
• Chain reasoning is the logical pattern for diagramming the if/then argument used to build the case for a literature review. • The first argument is constructed citing certain facts as evidence (R1 . . . . Rn), logically leading to the conclusions (C1). • The conclusions drawn from the first argument become the evidence (C1) for the second argument. • Using the evidence (C1) as their bases, conclusions (C2) are drawn to directly address the question posed, thereby creating the thesis (T).
– Is the argument pattern type legitimate? – Have the prerequisite conditions associated with the argument type been satisfied?
• Each of the argument pattern types has a rule of logic that makes it operable. Each relies on this rule as a specific condition that you must fulfill to use the argument pattern type correctly. • Without meeting these conditions, the rule of logic that forms the argument pattern type is not valid.
人脸识别文献
人脸识别文献人脸识别技术在当今社会中得到了广泛的应用，其应用领域涵盖了安全监控、人脸支付、人脸解锁等多个领域。
为了了解人脸识别技术的发展，下面就展示一些相关的参考文献。
1. 《Face Recognition: A Literature Survey》- 作者: Rabia Jafri, Shehzad Tanveer, and Mubashir Ahmad这篇综述性文献回顾了人脸识别领域的相关研究，包括了人脸检测、特征提取、特征匹配以及人脸识别系统的性能评估等。
该文中给出了对不同方法的综合评估，如传统的基于统计、线性判别分析以及近年来基于深度学习的方法。
2. 《Deep Face Recognition: A Survey》- 作者: Mei Wang, Weihong Deng该综述性文献聚焦于深度学习在人脸识别中的应用。
文中详细介绍了深度学习中的卷积神经网络（Convolutional Neural Networks, CNN）以及其在人脸特征学习和人脸识别中的应用。
同时，文中还回顾了一些具有代表性的深度学习人脸识别方法，如DeepFace、VGG-Face以及FaceNet。
3. 《A Survey on Face Recognition: Advances and Challenges》-作者: Anil K. Jain, Arun Ross, and Prabhakar这篇综述性文献回顾了人脸识别技术中的进展和挑战。
文中首先介绍了人脸识别技术的基本概念和流程，然后综述了传统的人脸识别方法和基于机器学习的方法。
此外，该文还介绍了一些面部表情识别、年龄识别和性别识别等相关技术。
4. 《Face Recognition Across Age Progression: A Comprehensive Survey》- 作者: Weihong Deng, Jiani Hu, Jun Guo该综述性文献主要关注跨年龄变化的人脸识别问题。
英文文献综述标准范文
英⽂⽂献综述标准范⽂ 下⾯是店铺为⼤家整理的⼀些关于“英⽂⽂献综述标准范⽂”的资料，供⼤家参阅。
英⽂⽂献综述范⽂ How to Write a Literature Review ? I. The definition of Literature Review ⽂献综述(Literautre Review)是科研论⽂中重要的⽂体之⼀。
它以作者对各种⽂献资料的整理、归纳、分析和⽐较为基础，就某个专题的历史背景、前⼈的⼯作、研究现状、争论的焦点及发展前景等⽅⾯进⾏综合、总结和评论。
通过阅读⽂献综述，科研⼯作者可花费较少的时间获得较多的关于某⼀专题系统⽽具体的信息，了解其研究现状、存在的问题和未来的发展⽅向。
II. The purposes of literature review And Its Components A. The Purposes On the one hand, it helps you broaden the view and perspective of the topic for your graduation thesis. On the other hand, it helps you narrow down the topic and arrive at a focused research question. B. Its Components There are six parts in a complete Literature Review. 标题与作者(title and author) 摘要与关键词(abstract and key words) 引⾔(introduction) 述评(review) 结论(conclusion) 参考⽂献(references) III. Classification of Source Materials How can we locate the materials relevant to our topics better and faster? Basically, all these source materials may be classified into four majors of sources. A: Background sources: Basic information which can usually be found in dictionaries and encyclopedia complied by major scholars or founders of the field. Three very good and commonly recommenced encyclopedias are encyclopedias ABC, namely, Encyclopedia Americana, Encyclopedia Britannica, and Collier’s Encyclopedia. There are also reference works more specialized, such as The Encyclopedia of Language and Linguistics for linguistics and TEFL studies. Moreover, you may also find Encyclopedia on the web. B: Primary sources Those providing direct evidence, such as works of scholars of the field, biographies or autobiographies, memoirs, speeches, lectures, diaries, collection of letters, interviews, case studies, approaches, etc. Primary sources come in various shapes and sizes, and often you have to do a little bit of research about the source to make sure you have correctly identified it. When a first search yields too few results, try searching by broader topic; when a search yields too many results, refine your search by narrowing down your search. C: Secondary sources Those providing indirect evidence, such as research articles or papers, book reviews, assays, journal articles by experts in a given field, studies on authors or writers and their works, etc. Secondary sources will inform most of your writing in college. You will often be asked to research your topic using primary sources, but secondary sources will tell you which primary sources you should use and will help you interpret those primary sources. To use theme well, however, you need to think critically them. There are two parts of a source that you need to analyze: the text itself and the argument within the text. D: Web sources The sources or information from websites. Web serves as an excellent resource for your materials. However, you need to select and evaluate Web sources with special care for very often Web sources lack quality control. You may start with search engines, such as Google, Yahoo, Ask, Excite, etc. It’s a good idea to try more than one search engine, since each locates sources in its own way. When using websites for information, be sure to take care for the authorship and sponsorship. If they are both unclear, be critical when you use information. The currency of website information should also be taken into account. Don’t use too out information dated for your purpose. IV. Major strategies of Selecting Materials for literature review A. Choosing primary sources rather than secondary sources If you have two sources, one of them summarizing or explaining a work and the other the work itself, choose the work itself. Never attempt to write a paper on a topic without reading the original source. B. Choosing sources that give a variety of viewpoints on your thesis Remember that good argument essays take into account counter arguments. Do not reject a source because it makes an argument against you thesis. C. Choosing sources that cover the topic in depth Probably most books on Communicative Language Teaching mention William Littlewood, but if this your topic, you will find that few sources cover the topic in depth. Choose those. D. Choosing sources written by acknowledged experts If you have a choice between an article written by a freelance journalist on Task-based Teaching and one written by a recognized expert like David Nunan, Choose the article by the expert. E. Choosing the most current sources If your topic involves a current issue or social problem or development in a scientific field, it is essential to find the latest possible information. If all the books on these topics are rather old, you probably need to look for information in periodicals. V. Writing a literature Review A. When you review related literature, the major review focuses should be: 1. The prevailing and current theories which underlie the research problem. 2. The main controversies about the issue, and about the problem. 3. The major findings in the area, by whom and when. 4. The studies which can be considered the better ones, and why. 5. Description of the types of research studies which can provide the basis for the current theories and controversies. 6. Criticism of the work in the area. B. When you write literature review, the two principles to follow are: 1. Review the sources that are most relevant to your to your thesis. 2. Describe or write your review as clear and objective as you can. C. Some tips for writing the review: 1. Define key terms or concepts clearly and relevant to your topic. 2. Discuss the least-related references to your question first and the most related references last. 3. Conclude your review with a brief summary. 4. Start writing your review early. VI. ⽂献综述主要部分的细节性提⽰和注意事项 主要部分细节提⽰： 引⾔(Introduction) 引⾔是⽂献综述正⽂的开始部分，主要包括两个内容：⼀是提出问题;⼆是介绍综述的范围 和内容。
如何阅读文献——文献阅读三步法paper-reading
How to Read a PaperAugust2,2013S.KeshavDavid R.Cheriton School of Computer Science,University of WaterlooWaterloo,ON,Canadakeshav@uwaterloo.caABSTRACTResearchers spend a great deal of time reading research pa-pers.However,this skill is rarely taught,leading to much wasted effort.This article outlines a practical and efficient three-pass method for reading research papers.I also de-scribe how to use this method to do a literature survey. 1.INTRODUCTIONResearchers must read papers for several reasons:to re-view them for a conference or a class,to keep current in theirfield,or for a literature survey of a newfield.A typi-cal researcher will likely spend hundreds of hours every year reading papers.Learning to efficiently read a paper is a critical but rarely taught skill.Beginning graduate students,therefore,must learn on their own using trial and error.Students waste much effort in the process and are frequently driven to frus-tration.For many years I have used a simple‘three-pass’approach to prevent me from drowning in the details of a paper be-fore getting a bird’s-eye-view.It allows me to estimate the amount of time required to review a set of papers.Moreover, I can adjust the depth of paper evaluation depending on my needs and how much time I have.This paper describes the approach and its use in doing a literature survey.2.THE THREE-PASS APPROACHThe key idea is that you should read the paper in up to three passes,instead of starting at the beginning and plow-ing your way to the end.Each pass accomplishes specific goals and builds upon the previous pass:The first pass gives you a general idea about the paper.The second pass lets you grasp the paper’s content,but not its details.The third pass helps you understand the paper in depth.2.1Thefirst passThefirst pass is a quick scan to get a bird’s-eye view of the paper.You can also decide whether you need to do any more passes.This pass should take aboutfive to ten minutes and consists of the following steps:1.Carefully read the title,abstract,and introduction2.Read the section and sub-section headings,but ignoreeverything else3.Glance at the mathematical content(if any)to deter-mine the underlying theoretical foundations4.Read the conclusions5.Glance over the references,mentally ticking offtheones you’ve already readAt the end of thefirst pass,you should be able to answer thefive Cs:1.Category:What type of paper is this?A measure-ment paper?An analysis of an existing system?A description of a research prototype?2.Context:Which other papers is it related to?Whichtheoretical bases were used to analyze the problem?3.Correctness:Do the assumptions appear to be valid?4.Contributions:What are the paper’s main contribu-tions?5.Clarity:Is the paper well written?Using this information,you may choose not to read fur-ther(and not print it out,thus saving trees).This could be because the paper doesn’t interest you,or you don’t know enough about the area to understand the paper,or that the authors make invalid assumptions.Thefirst pass is ade-quate for papers that aren’t in your research area,but may someday prove relevant.Incidentally,when you write a paper,you can expect most reviewers(and readers)to make only one pass over it.Take care to choose coherent section and sub-section titles and to write concise and comprehensive abstracts.If a reviewer cannot understand the gist after one pass,the paper will likely be rejected;if a reader cannot understand the high-lights of the paper afterfive minutes,the paper will likely never be read.For these reasons,a‘graphical abstract’that summarizes a paper with a single well-chosenfigure is an ex-cellent idea and can be increasingly found in scientific jour-nals.2.2The second passIn the second pass,read the paper with greater care,but ignore details such as proofs.It helps to jot down the key points,or to make comments in the margins,as you read. Dominik Grusemann from Uni Augsburg suggests that you “note down terms you didn’t understand,or questions you may want to ask the author.”If you are acting as a paper referee,these comments will help you when you are writing your review,and to back up your review during the program committee meeting.1.Look carefully at thefigures,diagrams and other illus-trations in the paper.Pay special attention to graphs.Are the axes properly labeled?Are results shown witherror bars,so that conclusions are statistically sig-nificant?Common mistakes like these will separaterushed,shoddy work from the truly excellent.2.Remember to mark relevant unread references for fur-ther reading(this is a good way to learn more aboutthe background of the paper).The second pass should take up to an hour for an expe-rienced reader.After this pass,you should be able to grasp the content of the paper.You should be able to summarize the main thrust of the paper,with supporting evidence,to someone else.This level of detail is appropriate for a paper in which you are interested,but does not lie in your research speciality.Sometimes you won’t understand a paper even at the end of the second pass.This may be because the subject matter is new to you,with unfamiliar terminology and acronyms. Or the authors may use a proof or experimental technique that you don’t understand,so that the bulk of the pa-per is incomprehensible.The paper may be poorly written with unsubstantiated assertions and numerous forward ref-erences.Or it could just be that it’s late at night and you’re tired.You can now choose to:(a)set the paper aside,hoping you don’t need to understand the material to be successful in your career,(b)return to the paper later,perhaps after reading background material or(c)persevere and go on to the third pass.2.3The third passTo fully understand a paper,particularly if you are re-viewer,requires a third pass.The key to the third pass is to attempt to virtually re-implement the paper:that is, making the same assumptions as the authors,re-create the work.By comparing this re-creation with the actual paper, you can easily identify not only a paper’s innovations,but also its hidden failings and assumptions.This pass requires great attention to detail.You should identify and challenge every assumption in every statement. Moreover,you should think about how you yourself would present a particular idea.This comparison of the actual with the virtual lends a sharp insight into the proof and presentation techniques in the paper and you can very likely add this to your repertoire of tools.During this pass,you should also jot down ideas for future work.This pass can take many hours for beginners and more than an hour or two even for an experienced reader.At the end of this pass,you should be able to reconstruct the entire structure of the paper from memory,as well as be able to identify its strong and weak points.In particular,you should be able to pinpoint implicit assumptions,missing citations to relevant work,and potential issues with experimental or analytical techniques.3.DOING A LITERATURE SURVEYPaper reading skills are put to the test in doing a literature survey.This will require you to read tens of papers,perhaps in an unfamiliarfield.What papers should you read?Here is how you can use the three-pass approach to help.First,use an academic search engine such as Google Scholar or CiteSeer and some well-chosen keywords tofind three to five recent highly-cited papers in the area.Do one pass on each paper to get a sense of the work,then read their re-lated work sections.You willfind a thumbnail summary ofthe recent work,and perhaps,if you are lucky,a pointer toa recent survey paper.If you canfind such a survey,youare done.Read the survey,congratulating yourself on your good luck.Otherwise,in the second step,find shared citations and repeated author names in the bibliography.These are thekey papers and researchers in that area.Download the key papers and set them aside.Then go to the websites of thekey researchers and see where they’ve published recently. That will help you identify the top conferences in thatfield because the best researchers usually publish in the top con-ferences.The third step is to go to the website for these top con-ferences and look through their recent proceedings.A quick scan will usually identify recent high-quality related work. These papers,along with the ones you set aside earlier,con-stitute thefirst version of your survey.Make two passes through these papers.If they all cite a key paper that youdid notfind earlier,obtain and read it,iterating as neces-sary.4.RELATED WORKIf you are reading a paper to do a review,you should also read Timothy Roscoe’s paper on“Writing reviews for sys-tems conferences”[3].If you’re planning to write a technical paper,you should refer both to Henning Schulzrinne’s com-prehensive web site[4]and George Whitesides’s excellent overview of the process[5].Finally,Simon Peyton Joneshas a website that covers the entire spectrum of research skills[2].Iain H.McLean of Psychology,Inc.has put together a downloadable‘review matrix’that simplifies paper review-ing using the three-pass approach for papers in experimen-tal psychology[1],which can probably be used,with minor modifications,for papers in other areas.5.ACKNOWLEDGMENTSThefirst version of this document was drafted by my stu-dents:Hossein Falaki,Earl Oliver,and Sumair Ur Rahman.My thanks to them.I also benefited from Christophe Diot’s perceptive comments and Nicole Keshav’s eagle-eyed copy-editing.I would like to make this a living document,updating itas I receive comments.Please take a moment to email meany comments or suggestions for improvement.Thanks to encouraging feedback from many correspondents over the years.6.REFERENCES[1]I.H.McLean,“Literature Review Matrix,”/[2]S.Peyton Jones,“Research Skills,”/en-us/um/people/simonpj/papers/giving-a-talk/giving-a-talk.htm[3]T.Roscoe,“Writing Reviews for Systems Conferences,”http://people.inf.ethz.ch/troscoe/pubs/review-writing.pdf[4]H.Schulzrinne,“Writing Technical Articles,”/∼hgs/etc/writing-style.html[5]G.M.Whitesides,“Whitesides’Group:Writing a Paper,”/∼rlake/Whitesides writing res paper.pdf。
如何写好literature review survey
Writing a Literature ReviewA literature review…•Provides an overview and a critical evaluation of a body of literature relating to a research topic ora research problem.•Analyzes a body of literature in order to classify it by themes or categories, rather than simply discussing individual works one after another.•Presents the research and ideas of the field rather than each individual work or author by itself.A literature review often forms part of a larger research project, such as within a thesis (or major research paper), or it may be an independent written work, such as a synthesis paper.Purpose of a literature reviewA literature review situates your topic in relation to previous research and illuminates a spot for your research. It accomplishes several goals:•provides background for your topic using previous research.•shows you are familiar with previous, relevant research.•evaluates the depth and breadth of the research in regards to your topic.•determines remaining questions or aspects of your topic in need of research.Relationship between a literature review and a research projectAcademic research at the graduate level is always part of a dialogue among researchers. As a graduate student, you must therefore indicate that you know where your topic is positioned within your field of study.Therefore, a literature review is a key part of most research projects at the graduate level. There is often a reciprocal relationship between a literature review and the research project for which it is written:• A research project is often undertaken in response to a literature review. Doing the literature review for a topic often reveals areas requiring further research. In this way, writing the literature review helps to formulate the research question.• A literature review helps to establish the validity of a research project by revealing gaps in the existing literature on a topic that offer opportunities for new research.Importance of the research questionOnce identified, the research question will drive the research project. Whatever you read or write should have a clear connection to your question.How to write a strong literature reviewThere are several steps toward writing a strong literature review:1.Synthesize and evaluate information2.Identify the main ideas of the literature3.Identify the main argument of the literature reviewanize the main points of the literature review5.Write literature review1. Synthesize and evaluate informationA literature review requires critical thinking, reading, and writing. You will take the information that you have gathered through your research and synthesize and evaluate it by indicating important ideas and trends in the literature and explaining their significance.Strategies for reading•As soon as you begin reading, take note of the themes or categories that you see emerging. These may be used later to develop a structure for the literature review.•Take note of how other writers classify their data, the literature in their fields, etc. It can be helpful to read literature reviews in your discipline to see how they are structured.Categories for analysis and comparisonA strong literature review examines each work on its own and in relation to other works by identifying and then analyzing them with regards to a number of different research aspects and ideas. Here are some possible categories to use for comparison and analysis.topicargumentresults found and conclusions methodstheoretical approach key wordsOverall, a literature review seeks to answer the following questions:•What does the literature tell you?•What does the literature not tell you?•Why is this important?Questions for analyzing individual works-What is the argument? Is it logically developed? Is it well defended?-What kind of research is presented? What are the methods used? Do they allow the author to address your research question effectively? Is each argument or point based on relevant research?If not, why?-What theoretical approach does the author adopt? Does it allow the researcher to make convincing points and draw convincing conclusions? Are the author’s biases and presuppositions openlypresented, or do you have to identify them indirectly? If so, why?-Overall, how convincing is the argument? Are the conclusions relevant to the field of study? Questions for comparing works-What are the main arguments? Do the authors make similar or different arguments? Are some arguments more convincing than others?-How has research been conducted in the literature? How extensive has it been? What kinds of datahave been presented? How pertinent are they? Are there sufficient amounts of data? Do theyadequately answer the questions?-What are the different types of methodologies used? How well do they work? Is one methodology more effective than others? Why?-What are the different theoretical frameworks or approaches used? What do they allow the authors to do? How well do they work? Is one approach more effective than others? Why?-Overall, is one work more convincing than others? Why? Or are the works you have compared too different to evaluate against each other?The Academic Writing Help Centre offers more information on synthesis and evaluation in the discussion group and accompanying handout on Information Management for a Literature Review.2. Identify the main ideas of the literatureOnce you have begun to synthesize your research, you will begin to identify some main ideas and trends that pervade the topic or that pertain to your research question.Use these main ideas to classify the information and sources that you have read. Later, these ideas can be used as the main topics of discussion in the literature review, and if you have already organized your literature on these topics, it will be easy to summarize the literature, find examples, etc.3. Identify the main argument of the literature reviewJust like any academic paper, the literature review should have a main idea about the literature that you would like the readers to understand. This argument is closely related to your research question in that it presents a situation in the body of literature which motivates your research question.ExampleArgument from a literature review: “Although some historians make a correlation between the Ukrainian Catholic and Orthodox churches and the retention of Ukrainian culture and language by Ukrainian immigrants in Canada, little has been said of the role of the Roman Catholic Church in the development of Ukrainian communities in Canada.”Research question:“How has the Roman Catholic Church shaped Ukrainian-Canadian identity?”4. Organize the main points of the literature reviewAfter identifying the main ideas that need to be presented in the literature review, you will organize them in such a way as to support the main argument. A well-organized literature review presents the relevant aspects of the topic in a coherent order that leads readers to understand the context and significance of your research question and project.As you organize the ideas for writing, keep track of the supporting ideas, examples, and sources that you will be using for each point.5. Write the literature reviewOnce the main ideas of the literature review are in order, writing can flow much more smoothly. The following tips provide some strategies to make your literature review even stronger.Tips for Writing and PresentationGive structure to the literature review.Like any academic paper, a literature review should contain an introduction, a body and a conclusion, and should be centered on a main idea or argument about the literature you are reviewing.If the literature review is a longer document or section, section headers can be useful to highlight the main points for the reader. However, the different sections should still flow together.Explain the relevance of material you use and cite.It is important to show that you know what other authors have written on your topic. However, you should not simply restate what others have said; rather, explain what the information or quoted material means in relation to your literature review.•Is there a relevant connection between a specific quote or information and the corresponding argument or point you are making about the literature? What is it?•Why is it necessary to include this piece of information or quote?Use verb tenses strategically.•Present tense is used for relating what other authors say and for discussing the literature, theoretical concepts, methods, etc.“In her article on biodiversity, Jones stipulates that ….”In addition, use the present tense when you present your observations on the literature.“However, on the important question of extinction, Jones remains silent.”•Past tense is used for recounting events, results found, etc.“Jones and Green conducted experiments over a ten-year period. They determined that it was not possible to recreate the specimen.”BibliographyBell, Judith. Doing Your Research Project: A Guide for First-time Researchers in Education, Health and Social Science.Maidenhead, Berkshire: Open University Press, 2005.Boote, David N. and Penny Beile. “Scholars before researchers: On the centrality of the dissertation literature review in research preparation.” Educational researcher, 34.6 (2005): 3-15.Booth, Wayne C., Gregory G. Colomb, and Joseph M. Williams. The Craft of Research. Chicago: University of Chicago Press, 2003.Verma, Gajendra K. and Kanka Mallick. Researching Education: Perspectives and Techniques. London: Falmer Press, 1999.The Writing Center, University of North Carolina – Chapel Hill. Literature Reviews. Chapel Hill, NC. 2005. Available /depts/wcweb/handouts/literature_review.html.© 2007 Academic Writing Help Centre, University of Ottawawww.sass.uottawa.ca/writing 613-562-5601 cartu@uOttawa.ca。
Structure of reports 报告的结构
Structure of reports 报告的结构TitleThe title needs to concisely state the topic of the report. It needs to be informative and descriptives so that someone just reading the title will understand the main issue of your report. You don‟t need to include excessive detail in your title but avoid being vague and too general.>> back to topAbstract 摘要(Also called the Summary or Executive Summary)This is the …shop window‟ for your report. It is the first (and sometimes the only) section to be read and should be the last to be written. It should enable the reader to make an informed decision about whether they want to read the whole report. The length will depend on the extent of the work reported but it is usually a paragraph or two and always less than a page.A good way to write an abstract is to think of it as a series of brief answers to questions. These would probably include:* What is the purpose of the work?* What methods did you use for your research?* What were the main findings and conclusions reached as a result of your research?* Did your work lead you to make any recommendations for future actions?>> back to topIntroduction(Also called Background or Context)In this section you explain the rationale for undertaking the work reported on, including what you have been asked (or chosen) to do, the reasons for doing it and the background to the study. It should be written in an explanatory style.State what the report is about - what is the question you are trying to answer? If it is a brief for a specific reader (e.g. a feasibility report on a construction project for a client), say who they are. Describe your starting point and the background to the subject, for instance: what research hasalready been done (if you have been asked to include a Literature Survey later in the report, you only need a brief outline of previous research in the Introduction); what are the relevant themes and issues; why are you being asked to investigate it now?Explain how you are going to go about responding to the brief. If you are going to test a hypothesis in your research, include this at the end of your introduction. Include a brief outline of your method of enquiry. State the limits of your research and reasons for them, for example; “Research will focus on native English speakers only, as a proper consideration of the issues arising from speaking English as a second language is beyond the scope of this project”.>> back to topLiterature Survey 文献调查，文献研究(Also called Literature Review or Survey/Review of Research)This is a survey of publications (books, journals, authoritative websites, sometimes conference papers) reporting work that has already been done on the topic of your report. It should only include studies that have direct relevance to your research.A literature survey should be written like an essay in a discursive style, with an introduction, main discussion grouped in themes and a conclusion.Introduce your review by explaining how you went about finding your materials, and any clear trends in research that have emerged. Group your texts in themes. Write about each theme as a separate section, giving a critical summary of each piece of work and showing its relevance to your research.Conclude with how the review has informed your research (things you‟ll be building on, gaps you‟ll be filling etc).>> back to topMethods(Also called Methodology)You need to write your Methods section in such a way that a reader could replicate the research you have done. There should be no ambiguity here, so you need to write in a very factual informative style.You need to state clearly how you carried out your investigation. Explain why you chose this particular method (questionnaires, focus group, experimental procedure etc), include techniques and any equipment you used. If there were participants in your research, who were they? How many? How were they selected?Write this section concisely but thoroughly – go through what you did step by step, including everything that is relevant. You know what you did, but could a reader follow your description?>> back to topResults(Also called Data or Findings)This section has only one job which is to present the findings of your research as simply and clearly as possible. Use the format that will achieve this most effectively e.g. text, graphs, tables or diagrams.When deciding on a graphical format to use, think about how the data will look to the reader. Choose just one format - don‟t repeat the same information in, for ins tance, a graph and a table. Label your graphs and tables clearly. Give each figure a title and describe in words what the figure demonstrates. For a beginners' guide to data analysis, see Analyse This!Writing in this section should be clear, factual and informative. Save your interpretation of the results for the Discussion section.>> back to topDiscussionThis is probably the longest section and worth spending time on. It brings everything together, showing how your findings respond to the brief you explained in your introduction and the previous research you surveyed in your literature survey. It should be written in a discursive style, meaning you need to discuss not only what your findings show but why they show this, using evidence from previous research to back up your explanations.This is also the place to mention if there were any problems (for instance, if your results were different from expectations, you couldn‟t find important data, or you had to change your method or participants) and how they were or could have been solved.>> back to topConclusionYour conclusions should be a short section with no new arguments or evidence. Sum up the main points of your research - how do they answer the original brief for the work reported on?This section may also include:* Recommendations for action* Suggestions for further research>> back to topReferences(Also called Reference List or Bibliography)List here full details for any works you have referred to in the report, including books, journals, websites and other materials. You may also need to list works you have used in preparing your report but have not explicitly referred to - check your instructions for this and for the correct style of referencing to use.You can find information about how to reference more unusual materials (television programmes, blogs etc) from various websites including the LearnHigher website on referencing. If you're not sure, the rule is to be consistent and to give enough details that a reader can find the same piece of information that you used.>> back to topAppendicesThe appendices hold any additional information that may help the reader but is not essential to the report‟s main findings: anything that 'adds value'. That might include (for instance) interview questions, raw data or a glossary of terms used. Label all appendices and refer to them where appropriate in the main text (e.g. …See Appendix A for an example questionnaire‟).>> back to topWhich section should I write first?It can be helpful to write up sections as you go along. This means that you write about what you've done while it's still fresh in your mind and you can see more easily if there are any gaps that might need additional research to fill them. In addition, you don't end up with a large piece of writing to do in one go - that can be overwhelming.Here is a suggested order for writing the main sections:1. Methods and Data/Results: As a rough guide, the more factual the section, the earlier youshould write it. So sections describing …what you did and what you found‟ a re likely to bewritten first.2. Introduction and Literature Survey: Sections that explain or expand on the purpose of theresearch should be next. What questions are you seeking to answer, how did they arise,why are they worth investigating? These will help you to see how to interpret and analyseyour findings.3. Discussion: Once you‟ve established the questions your research is seeking to answer, youwill be able to see how your results contribute to the answers and what kind of answersthey point to. Write this early enough that you still have time to fill any gaps you find.4. Conclusions and Recommendations: These should follow logically from your Discussion.They should state your conclusions and recommendations clearly and simply.5. Abstract/Executive Summary: Once the main body is finished you can write a succinct andaccurate summary of the main features.>> back to topMy re port doesn’t seem to fit into these sectionsIf you haven‟t been given instructions on how to structure your report, look at examples of other reports in your discipline. Your department may have examples of past report writing assignments that you can see. Or try the UniLearning website which has a useful guide to features of reports in various disciplines.For some reports, (often business or management reports) it isn‟t appropriate to use the 'in troduction, methods, results, discussion, conclusion' model. Instead, you have to create appropriatesub-headings depending on the brief you have been given.All reports aim to inform the reader about a specific investigation so you need to select the best headings to lead the reader through the different stages of this investigation. Read your brief carefully, brainstorm what you need to include, then group similar ideas together; see if these groups would make logical sub-headings.undefined undefined undefined undefined undefined undefinedYou are given the following brief:'Select a particular job in your chosen field and research what the role involves, what the career prospects are, which companies hire for this role and what skills are required.'How would you structure this report?A possible structure could be:1. Introduction: Background to the role, brief description of what the job involves and how to find information about it.2. Job description and skills: Detailed description of the responsibilities of the role and the skills required.3. Relevant employers: Which companies hire for this role and what they are looking for.4. Career prospects: What other jobs might this role lead onto, what is the job market like for this role?5. Conclusion: What this research has shown about the best ways of becoming employed in this role.>> back to top。
Survey_the_Literature
This scheme shows connections between groups of data. Here you examine likenesses and differences in each group by comparing and contrasting the evidence and claims. As with side-by-side reasoning, cite expert opinions, research studies, statistics, and expert testimony to build an evidentiary pattern for the first claim group (A). The set of data from the next claim group (B) is also presented. Look at the differences and likenesses between the data , and compare and contrast the two side-by-side arguments.
ห้องสมุดไป่ตู้
Serial in nature, it begins by citing one or more reasons that justify a conclusion. It uses a one-on-one reasoning pattern as its foundation. The conclusion of the first pattern then becomes the evidence for the second conclusion. This line of logic continues until the final conclusion has been warranted. Notice that this pattern forms as if you were making a daisy chain
The Assembly Line Balancing Problem
วิศวกรรมสารมข. ป ท34 ฉบับที2(133 - 140) มีนาคม–เมษายน 2550บทความวิชาการKKU Engineering Journal Vol.34No .2 (133 - 140) March – April2007The Assembly Line Balancing Problem :Review articles*Nuchsara Kriengkorakot1) and Nalin Pianthong1)1)Industrial Engineering Department, Faculty of Engineering, Ubon Rajathanee University 34190Email : ennuchkr@ubu.ac.thABSTRACTThe assembly line balancing problem (ALBP) consists of assigning tasks to an ordered sequence of stations such that the precedence relations among the tasks are satisfied and some performance measure is optimized. In this survey paper we give an up-to-date review and discuss thedevelopment of the classification of the assembly line balancing problem (ALBP) which has attractedattention of researchers and practitioners of research for almost half a century. We also present suggestions and summaries for future research directions of the literature review.Keywords : Assembly line balancing problem (ALBP), Literature review, Optimization problem* Original manuscript submitted: October 22, 2006 and Final manuscript received: January 15, 2007134 Nuchsara Kriengkorakot and Nalin PianthongIntroductionThe fundamental of line balancing problems is to assign the tasks to an ordered sequence of stations, such that the precedence relations are satisfied and some measurements of effectiveness are optimized. (e.g.minimize the balance delay or minimize the number of work stations; etc)The first published paper of the assembly line balancing problem (ALBP) was made by Salveson(1955) who suggested a linear programming solution. Since then, the topic of line balancing has been of great interest to researchers. However, since the ALB problem falls into the NP hard class of combinatorialoptimization problems (Gutjahr and Nemhauser, 1964), it has consistently developed the efficient algorithms for obtaining optimal solutions. Thus numerous research efforts have been directed towards the development ofcomputer-efficient approximation algorithms or heuristics (e.g. Kilbridge and Wester, 1961; Helgeson and Birnie, 1961; Hoffman, 1963; Mansoor, 1964; Arcus, 1966; Baybar, 1986a) and exact methods to solve the ALBproblems. (e.g. Jackson, 1956; Bowman, 1960; Van Assche and Herroelen, 1978; Mamoud, 1989; Hackman et al., 1989; Sarin et al., 1999)In addition, with the growth of knowledge on the ALB problem, review articles are necessary to organize and summarize the finding for the researchers and practitioners. In fact, several articles (e.g. Kilbridgeand Wester, 1962; Mastor, 1970; Johnson,1981; Talbot et al.,1986; Baybars, 1986b; Ghosh and Gagnon, 1989; Erel and Sarin, 2001) have reviewed the work published on this problem. Recently two articles by Scholl andBecker (2006); Becker and Scholl (2006) provide the state-of-the-art exact and heuristic solution procedures for simple assembly line balancing (SALB) and a survey on problems and methods in generalized assembly linebalancing (GALB) respectively. In this paper, we give an up-to-date review, discuss about the development of the classification of the ALBP and summary of the future research suggestions from the review articles.A Definition of Assembly Line Balancing (ALB)Basic problem of ALBAn assembly line consists of work stations k = 1,…., m usually arranged along a conveyor belt or a similar material handling equipment. The jobs are consecutively launched down the line and are moved fromstation to station. At each station, certain operations are repeatedly performed regarding the cycle time. In general, the line balancing problem consists of optimally balancing the assembly work among all stations withrespect to some objective. For this purpose, the total amount of work necessary to assemble a work piece (job) is split up into a set V = {1,…., n} of elementary operations named tasks. Performing a task j takes a task time t jand requires certain equipment of machines and/or skills of workers. The total workload necessary for assembling a work piece is measured by the sum of task times ∑t. These elements can be summarized by a precedence diagram. It contains a node for each task, node weights for the task times, arcs the direct and paths for the indirectprecedence constraints. Figure 1 shows a precedence diagram with n = 9 tasks having task times between 2-9 in time unit.Figure 1. Precedence diagramThe Assembly Line Balancing Problem : Review articles 135A feasible line balance, i.e. an assignment of tasks to stations has to ensure that no precedence relationship is violated. The set S k of tasks assigned to a station k constitutes its station load or work content, the cumulated task time t (S k) = ∑∈Sk j j t is called station time. When a fixed cycle time c is given (paced line), a line balance is feasible only if the station time of neither station exceeds c. In case of t (S k) < c , the station k has an idle time of c – t(S k) time unit in each cycle. For example in Figure 1, a feasible line balancewith cycle time c =11 and number of station , m = 5 stations is given by stations loads ; S1 ={1,3}, S2 ={2,4},S3 = {5,6}, S4 = {7,8} , S5 = {9} A usual objective consists in maximizing the line utilization which is measured by the line efficiency E as the productive fraction of the line’s total operating time and directly depends on the cycle time c and the number of stations m. In the most simple case, the line efficiency is defined ∑as follows : E = )t/(cm*Classification of ALBPIn this section, we provide characteristics of balancing problems considered in the literature and give some classification schemes (c.f., e.g., Ghosh and Gagnon, 1989; Scholl and Becker, 2006; Becker and Scholl, 2006)1 Ghosh and Gagnon (1989) classified the ALBP into four categories; as shown in Figure 2(1) Single Model Deterministic (SMD) (2) Single Model stochastic (SMS)(3) Multi/Mixed Model Deterministic (MMD) (4) Multi/Mixed Model stochastic (MMS)The SMD version of the ALB problem assumes dedicated, single model assembly lines where the task times are known deterministically and an efficiency criterion is to be optimized. This is the original and simplest form of the assembly line balancing problem (SALB). Introduce other restrictions or factors(e.g. parallel stations, zoning restrictions) into the model and the problem becomes the General Assembly Line Balancing Problem (GALB)SMS problem category introduces the concept of task-time variability. This is more realistic forThemanual assembly lines, where workers’ operation times are seldom constant. With the introduction of stochastictask times many other issues become relevant, such as station times exceeding the cycle time (and perhaps the production of defective or unfinished parts), pacing effects on workers’ operation times, station lengths, the sizeand location of inventory buffers, launch rates and allocation of line imbalances.136 Nuchsara Kriengkorakot and Nalin PianthongMMD problem formulation assumes deterministic task times, but introduces the concept of an Theassembly line producing multiple products. Multi-model lines assemble two or more products separately in batches. In mixed-model lines single units of different models can be introduced in any order or mix to the line. Multi-mixed model lines introduce various issues that are not present in the single-model case. Model selection, model sequencing and launching rate(s) and model lot sizes become more critical issues here than in the single model case.MMS problem perspective differs from its MMD counterpart in that stochastic times are allowed. TheHowever, these issues become more complex for the MMS problem because factors such as learning effects, worker skill level, job design and worker task time variability become more difficult to analyze because the line is frequently rebalanced for each model assembled.2 Scholl and Becker (2006); Becker and Scholl (2006)They have classified the main characteristics of assembly line balancing problems considered in their several constraints and different objectives as shown in Figure 3. It has illustrated the classification of assembly line balancing problems.Figure 3. classification of assembly line balancing problems(1) SALBP : The simple assembly line balancing problem is relevant for straight single product assembly lines where only precedence constraints between tasks are considered (for a survey see Scholl and Becker, 2006)- Type 1 (SALB-1) of this problem consists of assigning tasks to work stations such that the number of stations (m) is minimized for a given production rate (fixed cycle time, c).- Type 2 (SALBP-2) is to minimize cycle time (maximize the production rate) for a given number of stations (m).- Type E (SALBP-E) is the most general problem version maximizing the line efficiency (E) thereby simultaneously minimizing c and m considering their interrelationship.- Type F (SALBP-F) is a feasibility problem which is to establish whether or not a feasible line balance exists for a given combination of m and c.(2) GALBP : In the literature, all problem types which generalize or remove some assumptions of SALBP are called generalized assembly line balancing problems (GALBP). This class of problems (including UALBP and MALBP) is very large and contains all problem extensions that might be relevant in practice including equipment selection, processing alternatives, assignment restrictions etc. (for a survey see Becker and Scholl, 2006).- MALBP and MSP : Mixed model assembly lines produce several models of a basic product in an intermixed sequence. Besides the mixed model assembly line balancing problem (MALBP), which has to assigntasks to stations considering the different task times for the different models and find a number of stations and aThe Assembly Line Balancing Problem : Review articles 137 cycle time as well as a line balance such that a capacity- or even cost-oriented objective is optimized (cf. Scholl, 1999, Chapter 3.2.2). However, the problem is more difficult than in the single-model case, because the station times of the different models have to be smoothed for each station (horizontal balancing; cf. Merengo et al., 1999). The better this horizontal balancing works, the better solutions are possible in the connected short-term mixed model sequencing problem (MSP). MSP has to find a sequence of all model units to be produced such that inefficiencies (work overload, line stoppage, off-line repair etc.) are minimized.(e.g. Bard et al., 1992 and Scholl et al., 1998)- UALBP : The U-line balancing problem (UALBP) considers the case of U-shaped (single product) assembly lines, where stations are arranged within a narrow U. As a consequence, worker are allowed to work on either side of the U, i.e. on early and late tasks in the production process simultaneously. Therefore, modified precedence constraints have to be observed. By analogy with SALBP, different problem types can be distinguished. (cf. Miltenburg and Wijngaard, 1994; Urban , 1998; Scholl and Klein,1999; Erel et al., 2001) Objective criteria and methodological techniques Objective criteria1 Ghosh and Gagnon (1989)Various technical and economic objective criteria have been used in the ALB literature ,Ghosh and Gagnon (1989). As can be seen in Table 1, the seven technical criteria, which relate to throughput operational efficiency, have in aggregate been the more commonly employed (58:13). Within the technical category, minimizing the number of work stations has been the most chosen.relate to assembly line operating cost or profitability measures. Table 1 liststypicallyEconomiccriteriasix economic criteria found in the literature, all the economic criteria consider labor cost or labor idleness cost, the most popular criterion and the apparent trend is to include other cost such as product incompletions (Kottas and Lau, 1973), penalty costs (Dar-El, 1977) and inventory and set-up costs (Caruso, 1965). While the technical criteria have been the classical dominant choice, economic criteria have gained rapid attention of researcher since the mid-1970s.Frequency of useType of objective criteriaSMD SMS MMD MMS total1. Technical-min. the no. of work st.(given cycle time) 16 2 2 1 21 -min. the cycle time (given no. of ws.) 13 1 2 0 16 -min. the total Idle time along the line 9 0 3 0 12 -min. the balance delay 2 0 1 0 3-min. the overall facility or line length 0 0 2 0 2-min. the throughput time 0 0 1 0 1-min. the probability that one or more0 1 1 1 3 work stations will exceed the cycle timetotal40 4 12 2 582. Economic-min. the combined cost of labour,0 3 0 1 4ws. and product incompleteness-min. the labour cost/unit 3 1 0 0 4-min. the total penalty cost for a no.0 0 2 0 2of inefficiencies-min. the inventory, set up and idle0 0 1 0 1time cost-min. the total in-process inventory costs 0 1 0 0 1-max. the net profit 1 0 0 0 1total 4 5 3 1 13Table 1 : ALB objective criteria (Ghosh and Gagnon, 1989)138 Nuchsara Kriengkorakot and Nalin Pianthong2 Scholl (1999) (Chapter 1.3.10)From literature, he has mainly classified the objective criteria into two categories; capacity oriented goals and cost oriented goals.• Capacity oriented goals. Most commonly, ALBP contain the objective of maximizing the capacity utilizatio n of the line which is measured by the line efficiency (E). In case of deterministic operation times and single-model production, the line efficiency only depends on the cycle time and the number of stations. The simpler objectives are often considered :- minimize the number of stations for a given cycle time.- minimize the cycle time for a given number of stations.- minimizing the balance delay time (sum of idle times) and the balance delay (percentage of idle times) over all stations, etc.• Cost oriented goals. An important objective is to minimize the total costs of line including long-term investment costs and short-term operating costs. Assembly line production systems may concern the following cost categories (i.e. costs of machinery and tools, wage costs, material costs, etc. (for detail see Scholl, 1999, Chapter 1.3.10 pp.21-22 ) which are mainly influenced by the cycle time and the number of stations (cf.Deckro, 1989; Rosenberg and Ziegler, 1992; Amen, 2000 a, 2000b, 2001 and Scholl and Becker, 2003).In the literature, usually only one objective is used, while other goals are formulated as constraints. Only few references deal with multiple objective assembly line balancing problems, namely, Baybars (1985), Shtub and Dar-El (1990), Deckro and Rangachari (1990), Malakooti (1991,1994) and Malakooti and Kumar (1996).Methodological techniquesThe large combinational complexity of the ALB problem has resulted in enormous computational difficulties. To achieve optimal or at least acceptable solutions, various solution methodologies have been explored. These methods are organized and their use analyzed in Table 2.Frequency of useALB techniquesSMD SMS MMD MMS total1. Exact methods- Linear Programming (LP) 1 0 0 0 1- Integer Programming (IP) 7 0 1 0 8- Dynamic Programming (DP) 4 2 0 0 6- Goal Programming (GP) 1 0 0 0 1- Shortest-path tech. (SP) 2 0 2 0 4- Maximal-path tech. (MP) 1 0 0 0 1- Branch and bound (BB) 11 1 0 1 13total 342. Inexact methods or heuristic- Priority ranking and assignment 10 5 7 *2+sim 24- Tree search (heuristic BB) 8 1 0 0 9- Trade and transfer 1 2 1 0 4- Random sampling 3 1 0 0 4- Others ; task grouping, approximation tech. 5 3 1*2+sim11total 52Table 2 : ALB methodological techniques (Ghosh and Gagnon, 1989)* Refers to two research studies which combined priority ranking procedures with simulation programs.Despite recent advances in problem formulation and solution procedure efficiency, mathematical programming/network-based optimization techniques are still computationally prohibitive beyond limited problem dimensions. It is also unlikely that computational efficiency will progress sufficiently in the near term to allow the use of even the most efficient exact techniques to obtain optimal solutions to realistically sized GALB problems. Therefore, heuristic and metaheuristic techniques (e.g. Simulated Annealing (SA), Tabu Search (TS),The Assembly Line Balancing Problem : Review articles 139 Genetic Algorithm (GA) and Ant Colony Optimization (ACO), etc.) still remain the only computationally efficient and sufficiently flexible methodologies capable of addressing large-scale, real-world ALB situations, particularly for the multi/mixed model and GALBP categories.Conclusion and future researchIn the literature survey, it shows that research has made significant algorithm developments in solving simple problems (SALBP). Though SALBP is a class of NP-hard optimization problems, effective exact and heuristic algorithms are available which solve small and medium-size instances of problems. Nevertheless, further algorithmic improvement is necessary for solving large-scale problems.Recently, assembly line balancing research evolved towards formulating and solving generalized problem (GALBP) with different additional characteristics such as cost functions, paralleling, equipment selection, u-line layout and mixed-model production. In the literature survey on GALBP (cf. Becker and Scholl, 2006) shows that a lot of relevant problems have been identified and modeled but development of sophisticated solution procedures has just begun. Then, additional research is necessary to adopt state-of-the-art solution concepts like metaheuristics and highly developed algorithms for SALBP to the variety of GALBP. Moreover, standardized and realistic test beds are required for testing and comparing methodical enhancements.REFERENCESAmen, M., 2000a. An exact method for cost-oriented assembly line balancing. International Journal of Production Economics. Vol.64: pp.187-195.Amen, M., 2000b. Heuristic methods for cost-oriented assembly line balancing: A survey. International Journal of Production Economics. Vol.68: pp.1-14.Amen, M., 2001. Heuristic methods for cost-oriented assembly line balancing: A comparison on solution quality and computing time. International Journal of Production Economics. Vol.69: pp.255-264.Arcus A.L., 1966. COMSOAL: A Computer Method of Sequencing Operations for Assembly Lines.International Journal of Production Research. Vol. 4, No.4: pp.259-277.Bard, J.F., Dar-El, E., Shtub, A., 1992. An analytic framework for sequencing mixed model assembly lines.International Journal of Production Research. Vol. 30: pp.35-48.Baybars, I., 1985. On currently practiced formulations of the assembly line balancing problem. Journal of Operations Management. Vol. 5: pp.449-453.Baybar, I., 1986a. An Efficient Heuristic Method for the Simple Assembly Line Balancing Problem.International Journal of Production Research. Vol. 24, No.1: pp.149-166.Baybars, I., 1986b. A Survey of Exact Algorithms for the Simple Assembly Line Balancing Problem.Management Science. Vol.32, No. 8: pp.909-932.Becker, C. and A. Scholl., 2006. A Survey on Problems and Methods in Generalized Assembly Line Balancing. European Journal of Operational Research. Volume 168, Issue 3: pp. 694-715.Bowman, E.H., 1960. Assembly Line Balancing by Linear Programming. Operations Research.Vol.8,No.3: pp.385-389.Caruso, F.R., 1965. Assembly line balancing for improved profits. Automation. Volume 12, January.Dar-El, E.M., and Curry, S., 1977. Optimal mixed-model sequencing for balanced assembly Lines. Omega.Vol. 5, pp.333-341.Deckro, R.F., 1989. Balancing cycle time and workstations. IIE Transactions. Vol. 21: pp.106-111.Deckro, R.F. and Rangachari, S., 1990. A Goal approach to assembly line balancing. Computers & Operations Research. Vol. 17: pp.509-521.Erel, E. and S.C. Sarin., 1998. A Survey of the Assembly Line Balancing Procedures. Production Planning and Control. Vol.9, No.5: pp.414-434.Erel E., I. Sabuncuoglu and B.A. Aksu, 2001. Balancing of U-type Assembly Systems Using Simulated Annealing. International Journal of Production Research. Vol. 39, No.13: pp.3003-3015.Ghosh S. and R. J. Gagnon., 1989. A Comprehensive Literature Review and Analysis of the Design, Balancing and Scheduling of Assembly Systems. International Journal of Production Research.Vol. 27, No.4: pp.637-670.Gutjahr, A. L., Nemhauser, G.L., 1964. An algorithm for the line balancing problem. Management Science.Vol. 11, No.2: pp. 08-315.140 Nuchsara Kriengkorakot and Nalin PianthongHackman, S.T., M. J. Magazine and T. S. Wee., 1989. Fast, Effective Algorithms for Simple Assembly Line Balancing Problems. Operations Research. Vol.37,No.6: pp.916-924.Helgeson, W.P. and Birnie, D.P., 1961. Assembly Line Balancing Using the Ranked Positional Weight-Technique. The Journal of Industrial Engineering. Vol. 12, No. 6: pp.394-398.Hoffmann,T.R., 1963. Assembly Line Balancing with a Precedence Matrix. Management Science. Vol. 9, No.4: pp.551-562.Jackson, J.R., 1956. A Computing Procedure for a Line Balancing Problem. Management Science. Vol.2,No.3: pp.261-271.Johnson, R.V., 1981. Assembly Line Balancing Algorithms: Computation Comparisons. International Journal of Production Research. Vol. 19, No.3: pp.277-287.Kilbridge, M.D. and Wester, L., 1961. A Heuristic Method of Assembly Line Balancing. The Journal of Industrial Engineering. Vol.12, No. 4: pp.292-298.Kilbridge, M.D. and Wester, L., 1962. A review of analytical systems of line balancing. Operations Research. Vol.10,No.5: pp.626-638.Kottas, J.F., and Lua, H.S. 1973. A cost-oriented approach to stochastic line balancing. AIIE Transactions. Vol.5, No.2.Malakooti, B., 1991. A multiple criteria decision making approach for the assembly line balancing problem.International Journal of Production Research. Vol. 29: pp.1979-2001.Malakooti, B., 1994. Assembly line balancing with buffers by multiple criteria optimization. International Journal of Production Research. Vol. 32: pp.2159-2178.Malakooti, B. and Kumar, A., 1996. An expert system for solving multi-objective assembly line balancing problem. International Journal of Production Research. Vol.34: pp.2533-2552.Mamoud,K.I., 1989. A Generalised Assembly Line Balancing Algorithm. PhD. Dissertation, University of Bradford, UK. 489 pages.Mansoor, E.M., 1964. Assembly Line Balancing -An Improvement on the Ranked Positional Weight Technique. The Journal of Industrial Engineering . Vol. 15, No.2: pp.73-77.Mastor, A.A., 1970. An Experimental Investigation and Comparative Evaluation of Production Line Balancing Techniques. Management Science. Vol.16, No.11: pp.728-746.Merengo, C., Nava, F., Pozetti, A., 1999. Balancing and sequencing manual mixed-model assembly lines.International Journal of Production Research. Vol. 37: pp.2835-2860.Miltenburg, G.J. and Wijngaard, J., 1994. The U-line Line Balancing Problem. Management Sciences. Vol.40, No.10: pp.1378-1388.Rosenberg, O. and Ziegler, H., 1992. A comparison of heuristic algorithms for cost-oriented assembly line balancing. Zeit-schrift für Operations Research. Vol. 35: pp.477-495.Salveson, M.E.,1955. The Assembly Line Balancing Problem. The Journal of Industrial Engineering.Vol.6,No.3: pp.18-25.Sarin, S.C., E. Erel and E.M. Dar-El., 1999. A Methodology for Solving Single-Model, Stochastic Assembly Line Balancing Problem. Omega. Vol. 27: pp.525-535.Scholl, A., 1999. Balancing and Sequencing of Assembly Lines. 2 ed., Heidelberg : Physica-Verlag.Scholl, A. and Becker, C., 2003. A note on an exact method for cost-oriented assembly line balancing. Jenar Schriften zur Wirtschaftswissenschaft 22/2003, University of Jena.Scholl, A. and Becker, C., 2006. State-of-the-art Exact and Heuristic Solution Procedures for Simple Assembly Line Balancing. European Journal of Operational Research. Volume 168, Issue 3: pp.666-693.Scholl, A. and Klein, R., 1999. ULINO : Optimally balancing U-shaped JIT assembly lines. International Journal of Production Research. Vol. 37, No.4: pp.721-736.Scholl, A., Klein,R., Domschke, W., 1998. Pattern based vocabulary building for effectively sequencing mixed model assembly lines. Journal of Heuristics. Vol. 4: pp.359-381.Shtub, A. and Dar-El, E.M., 1990. An assembly chart oriented assembly line balancing approach.International Journal of Production Research. Vol. 28: pp.1137-1151.Talbot, F. B., J.H. Patterson and W.V. Gehrlein., 1986. A Comparative Evaluation of Heuristic Line Balancing Techniques. Management Science. Vol. 32, No.4: pp.430-454.Uban, T.L., 1998. Note. Optimal Balancing of U-Shaped Assembly Lines. Management Sciences. Vol. 44, No.5: pp.738-741.Van Assche, F. and Herroelen,W.S., 1978. An Optimal Procedure for the Single-Model Deterministic Assembly Line Balancing Problem. European Journal of Operations Research. Vol.3: pp. 142-149.。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
8
Introduction (Cont’d)
• Software Reuse Process (Library Metaphor) – Developing for Reuse • Developing reusable artifacts – Developing with Reuse • Retrieval, Assessment, and Adaptation
9
Introduction (Cont’d)
• Software Retrieval System (Library Metaphor)
Retrieval Engine
INTPUT
OUTPUT
Query
Artifact Repository
10
Introduction (Cont’d)
• Retrieval = Search • Search = {Search Goal, Search Space, Comparison Function} • A similarity metric is measure of the degree of similarity of a pair of objects. Its converse is distance metric. • In software retrieval,
SIMILARITY METRIC FOR UML MODELS
Raimi Ayinde Rufai MS Thesis Oral Defense Information & Computer Science Department, King Fahd University of Petroleum & Minerals Dhahran 31261, KSA 20 January 2003
13
Motivation Why a similarity metric? • Retrieval • Tracing • Asset Scoping • Asset Mining Why a similarity metric for the UML? • The UML is the most popular representation language for OO reusable artifacts.
14
Literature Survey
15
Approach
• Assumptions
– Artifacts within the same domain should be more similar to each other than to artifacts from a different domain. – An artifact has the greatest similarity to itself. – The more common parts are added to a pair of artifacts the greater the similarity. – The more common parts are removed from a pair of artifacts, the lower their similarity.
Metric C
17
Approach (Cont’d)
• Inconsistency Penalty
B1 ⊗ B2 B1 + B2
I =1−
where B1 and B2 are binary matrices representing the mappings according to metrics M 1 and M 2 respectively. Operators ⊗ and + are the XOR and OR logical operators.
– search goal = query – Search space = artifact (asset) repository – Comparison function = similarity metric
• Summary: A software similarity metric has application in software retrieval. But is this the only application of a software similarity metric?
12
Research Problem
• Given a pair of software artifacts, represented in the Unified Model Language (UML), how can we measure the degree of similarity of these artifacts to each other.
– Signature-Based
• Signature-Based Similarity Metric (SBSM)
– Relationship-Based
• Relationship-Based Similarity Metric (RBSM)
19
Semantic Distance Measures
18
Approach (Cont’d)
• Approaches for Matching Class Models – Semantic-Based
• Deep Semantic Similarity Metric (DSSM) • Shallow Semantic Similarity Metric (SSSM)
5
Introduction (Cont’d)
• Reusable Artifacts include:
– – – – – – Domain Models Requirement Specifications Designs Documentation Test Data Source Code
6
Introduction (Cont’d)
• Paradigms of Software Reuse: – The Factory Metaphor – The Library Metaphor
7
Introduction (Cont’d)
• Software Products Lines (The Factory Metaphor)
– A set of software intensive systems that share a common, managed feature set satisfying a particular market segment’s specific needs or mission and that are developed from a common set of core assets in a predefined way – SEI. – Asset Scoping: Identifying the assets that should be made reusable [Harsu, 2001] – Asset Mining: Migrating legacy assets to product line [Eisenbarth and Simon, 2001]
– Structural view – Behavioral view – Use-case view
• The UML is a standard graphical language for visualizing, specifying, constructing and documenting a software intensive system – OMG • The UML has diagrams for each these views of software.
2
Introduction
• Software reuse is a process whereby an institution defines a set of systematic operating procedures to produce, classify and retrieve software artifacts for the purpose of using them in its development activities.
4
Introduction (Cont’d)
• Problems of software reuse include: increased maintenance costs, the NIH syndrome, lack of tool support, difficulty of maintaining a library of reusable artifacts, and the cost of locating and adapting reusable artifacts.
1
Agenda
• • • • • • • • • • • • Introduction System Viewpoints Research Problem Motivation Literature Survey Approach Proposed Metrics Validation Findings Contributions Further Work Questions
• Semantic Distance Measures
– Shallow Semantic Similarity Metric (SSSM) – Deep 20
Proposed Metrics – SSSM
1 if max( M 1 , M 2 ) = 0 SSSM ( M 1 , M 2 ) = 1 hso( name(Ci ( M 1 )), max if max( M 1 , M 2 ) > 0 max( M , M ) ∑ j name ( C ( M ))) i j 2 1 2