摘要
协同过滤推荐系统能够依据建立的用户概貌，过滤出用户感兴趣的信息并主动推荐给用户，可以有效解决互联网上出现的“信息过载”问题，已经被广泛应用在电子商务等诸多领域。然而，由于协同过滤推荐系统自身所具有的开放性，攻击者出于商业竞争等目的，人为地向系统注入大量虚假的用户概貌，企图使系统产生对他们有利的推荐结果。这种“托”攻击或推荐攻击给协同过滤推荐系统带来了极大的安全隐患。为了消除推荐攻击产生的安全隐患，关于推荐攻击检测方法的研究受到广泛关注。本文在对国内外研究现状综合分析的基础上，进一步对推荐攻击特征提取及检测方法进行了深入探讨。
首先，针对已有专用特征提取方法不能有效描述已知类型推荐攻击的问题，通过引入Hilbert-Huang变换、词频-逆向文档频率和互信息，提出一种推荐攻击专用特征提取方法。在分析已知类型推荐攻击的基础上，利用Hilbert-Huang变换、词频-逆向文档频率和互信息，提取已知类型推荐攻击的专用特征，作为检测已知类型推荐攻击的基础。
其次，针对已有通用特征提取方法不能有效描述未知类型推荐攻击的问题，通过引入信息熵，提出一种推荐攻击通用特征提取方法。从用户评分分布的角度，利用信息熵提取未知类型推荐攻击的通用特征，作为检测未知类型推荐攻击的基础。
再次，针对已有有监督检测方法误报率太高的问题，提出一种基于支持向量机的推荐攻击集成检测方法。利用上述提出的专用特征提取方法提取用户概貌的特征，利用随机采样技术生成有差异的基训练集，利用生成的基训练集训练支持向量机生成基分类器，对测试数据进行检测，采用多数投票机制融合基分类器的检测结果。
然后，针对已有检测方法不能有效检测未知推荐攻击的问题，提出一种基于仿生模式识别的未知推荐攻击集成检测方法，利用上述提出的通用特征提取方法提取用户概貌的特征，利用仿生模式识别技术覆盖真实概貌样本，将覆盖范围之外的用户概貌判断为攻击概貌，在此基础上，通过调整覆盖范围的大小生成基分类器，检测测试数据，采用多数投票机制融合基分类器的检测结果。
最后，在MovieLens数据集上与相关工作进行了实验对比，验证了所提方法的有效性。
Collaborative filtering recommender systems can filter out the information to satisfythe users' interests according to the established user profiles and recommend theinformation to users actively. They can solve the information overload problem on theInternet effectively, which have been widely used in many fields, e.g., e-commerce sites.Due to their natural openness, however, attackers artificially inject a large number of fakeprofiles into a collaborative filtering recommender system in order to bias therecommendation results to their advantage. These "shilling" attacks or recommendationattacks bring great security risk to collaborative recommender systems. To reduce thesecurity risk produced by recommendation attacks, the detection approaches forrecommendation attacks have attracted widespread attention. On the basis ofcomprehensive analysis for the current research in this area, this paper has conductedfurther deep research on the feature extraction methods and detection approaches forrecommendation attacks.
Firstly, aiming at the problem that the existing special feature extraction methods cannot describe the known recommendation attacks effectively, through introducingHilbert-Huang transform, term frequency-inverse document frequency, and mutualinformation a special feature extraction method for the known recommendation attacks isproposed. Based on the analysis of known recommendation attacks, Hilbert-Huangtransform, term frequency-inverse document frequency, and mutual information are usedto extract special features for these attacks. The extracted special features are used as thebasis of detecting known recommendation attacks.
Then, aiming at the problem that the existing general feature extraction methods cannot describe the unknown recommendation attacks effectively, through introducingentropy a general feature extraction method for the unknown recommendation attacks isproposed. From the perspective of user rating distribution, entropy is used to extractgeneral features for the unknown recommendation attacks. The extracted general featuresare used as the basis of detecting unknown recommendation attacks.
Next, aiming at the problem that the existing supervised detection approaches sufferfrom high false alarm ratio, an ensemble detection approach based on support vectormachine is proposed. The above proposed special features extraction method is used toextract features of user profiles. The bootstrap technique is used to generate the diversebase training sets. The generated base training sets are used to train support vectormachine to generate the base classifiers. These classifiers are used to detect the test sets.The majority voting strategy is used to integrate the detection results of the baseclassifiers.
After that, aiming at the problem that the existing detection approaches can not detectthe unknown recommendation attacks effectively, an ensemble detection approach basedon bionic pattern recognition is proposed. The above proposed general features extractionmethod is used to extract features of user profiles. The technique of bionic patternrecognition is used to cover the samples of genuine profiles. User profiles outside thecoverage are judged as attack profiles. On this basis, through adjusting the area of thecoverage the base classifiers are generated for the detection of test data. The majorityvoting strategy is used to integrate the detection results of the base classifiers.
Finally, the comparative experiments are conducted with the related work onMovieLens dataset. The effectiveness of the proposed approaches is verified.
引文
[1] Ngo-Ye T L, Sinha A P. Analyzing Online Review Helpfulness Using a RegressionalReliefF-Enhanced Text Mining Method[J]. ACM Transactions on Management InformationSystems,2012,3(2):1-20.
[2]贾大文,曾承,彭智勇,等.一种基于用户偏好自动分类的社会媒体共享和推荐方法[J].计算机学报,2012,35(11):2382-2391.
[3] Park Y J. The Adaptive Clustering Method for The Long Tail Problem of RecommenderSystems[J]. IEEE Transactions on Knowledge and Data Engineering,2013,25(8):1904-1915.
[4]孟祥武,胡勋,王立才,等.移动推荐系统及其应用[J].软件学报,2013,24(1):91-108.
[5] Gedikli F, Jannach D. Improving Recommendation Accuracy Based on Item-Specific TagPreferences[J]. ACM Transactions on Intelligent Systems and Technology,2013,4(1):1-19.
[6] Adomavicius G, Zhang J J. Stability of Recommendation Algorithms[J]. ACM Transactions onInformation Systems,2012,30(4):1-31.
[7] Biancalana C, Gasparetti F, Micarelli A, et al. An Approach to Social Recommendation forContext-Aware Mobile Services[J]. ACM Transactions on Intelligent Systems and Technology,2013,4(1):1-31.
[8] Adomavicius G, Kwon Y O. Improving Aggregate Recommendation Diversity UsingRanking-Based Techniques[J]. IEEE Transactions on Knowledge and Data Engineering,2012,24(5):896-911.
[9]刘建国,周涛,汪秉宏.个性化推荐系统的研究进展[J].自然科学进展,2009,19(1):1-15.
[10] Cached F, Carneiro V, Fernandez D, et al. Comparison of Collaborative Filtering Algorithms:Limitations of Current Techniques and Proposals for Scalable, High-Performance RecommenderSystems[J]. ACM Transactions on the Web,2011,5(1):1-23.
[11]赵琴琴,鲁凯,王斌. SPCF:一种基于内存的传播式协同过滤推荐算法[J].计算机学报,2013,36(3):671-676.
[12] Bellogin A, Cantador I, Diez Fernando, et al. An Empirical Comparison of Social CollaborativeFiltering and Hybrid Recommenders[J]. ACM Transactions on Intelligent Systems and Technology,2013,4(1):1-29.
[13] Victor P, Verbiest N, Cornelis C, et al. Enhancing the Trust-Based Recommendation Process withExplicit Distrust[J]. ACM Transactions on the Web,2013,7(2):1-19.
[14]吴湖,王永吉,王哲,等.两阶段联合聚类协同过滤算法[J].软件学报,2010,21(5):1042-1054.
[15]罗辛,欧阳元新,熊璋,等.通过相似度支持度优化基于K近邻的协同过滤算法[J].计算机学报,2010,33(8):1438-1445.
[16]黄创光,印鉴,汪静,等.不确定近邻的协同过滤推荐算法[J].计算机学报,2010,33(8):1370-1377.
[17] Bartolini I, Zhang Z J, Papadias D. Collaborative Filtering with Personalized Skylines[J]. IEEETransactions on Knowledge and Data Engineering,2011,23(2):190-203.
[18] Koren Y. Factor in the Neighbors: Scalable and Accurate Collaborative Filtering[J]. ACMTransactions on Knowledge Discovery From Data,2010,4(1):1-24.
[19] Lam S K, Riedl J. Shilling Recommender Systems for Fun and Profit[C]//Proceedings of the13thInternational Conference on World Wide Web. New York: ACM,2004:393-402.
[20] Aghili G, Shajari M, Khadivi S, et al. Using Genre Interest of Users to Detect Profile InjectionAttacks in Movie Recommender Systems[C]//Proceedings of the10th International Conference onMachine Learning and Applications. Washington: IEEE Computer Society,2011:49-52.
[21] Zhang S, Ouyang Y, Ford J, et al. Analysis of A Low-Dimensional Linear Model underRecommendation Attacks[C]//Proceedings of the29th Annual International ACM SIGIRConference on Research and Development in Information Retrieval. New York: ACM,2006:517-524.
[22] Mehta B, Hofmann T. A Survey of Attack-Resistant Collaborative Filtering Algorithms[J]. Bulletinof the Technical Committee on Data Engineering,2008,31(2):14-22.
[23] Burke R, Mobasher B, Williams C, et al. Classification Features for Attack Detection inCollaborative Recommender Systems[C]//Proceedings of the12th ACM SIGKDD InternationalConference on Knowledge Discovery and Data Mining. New York: ACM,2006:542-547.
[24] Hurley N, Cheng Z, Zhang M. Statistical Attack Detection[C]//Proceedings of the3rd ACMConference on Recommender Systems. New York: ACM,2009:149-156.
[25] Mobasher B, Burke R, Bhaumik R, et al. Towards Trustworthy Recommender Systems: AnAnalysis of Attack Models and Algorithm Robustness[J]. ACM Transactions on InternetTechnology,2007,7(4):1-38.
[26] Mobasher B, Burke R, Williams C, et al. Analysis and Detection of Segment-Focused AttacksAgainst Collaborative Recommendation[C]//Proceedings of the7th International Conference onKnowledge Discovery on the Web: Advances in Web Mining and Web Usage Analysis. Berlin:Springer,2006:96-118.
[27] Burke R, Mobasher B, Bhaumik R, et al. Segment-Based Injection Attacks against CollaborativeFiltering Recommender Systems[C]//Proceedings of the International Conference on Data Mining.Washington: IEEE Computer Society,2005:577-580.
[28] O’Mahony M P, Hurley N J, Silvestre G C M. Recommender Systems: Attack Types andStrategies[C]//Proceedings of the20th National Conference on Artificial Intelligence. Menlo Park,CA: AAAI,2005:334-339.
[29] Mobasher B, Burke R, Bhaumik R, et al. Attacks and Remedies in CollaborativeRecommendation[J]. IEEE Intelligent Systems,2007,22(3):56-63.
[30] Williams C, Mobasher B. Profile Injection Attack Detection for Securing CollaborativeRecommender Systems[R]. Chicago, Illinois: DePaul University,2006:1-47.
[31] Burke R, Mobasher B, Zabicki R, et al. Identifying Attack Models for SecureRecommendation[C]//Beyond Personalization: A Workshop on the Next Generation ofRecommender Systems Research. New York: ACM,2005:19-25.
[32] Mobasher B, Burke R, Bhaumik R, et al. Effective Attack Models for Shilling Item-BasedCollaborative Filtering System[C]//Proceedings of the7th International Workshop on KnowledgeDiscovery on the Web. Berlin: Springer,2005:13-23.
[33] Williams C, Mobasher B, Burke R, et al. Detection of Obfuscated Attacks in CollaborativeRecommender Systems[C]//Proceedings of the ECAI2006Workshop on Recommender Systems.Ohmsha: IOS,2006:19-23.
[34] Cheng Z P, Hurley N. Effective Diverse and Obfuscated Attacks on Model-Based RecommenderSystems[C]//Proceedings of the3rd ACM Conference on Recommender Systems. New York:ACM,2009:141-148.
[35] Mehta B, Nejdl W. Unsupervised Strategies for Shilling Detection and Robust CollaborativeFltering[J]. User Modeling and User-Adapted Interaction,2009,19(1-2):65-97.
[36] Chirita P A, Nejdl W, Zamfir C. Preventing Shilling Attacks in Online RecommenderSystems[C]//Proceedings of the7th Annual ACM International Workshop on Web Informationand Data Management. New York: ACM,2005:67-74.
[37] Su X F, Zeng H J, Chen Z. Finding Group Shilling in Recommendation System[C]//SpecialInterest Tracks and Posters of the14th International Conference on World Wide Web. New York:ACM,2005:960-961.
[38] Mehta B, Hofmann T, Fankhauser P. Lies and Propaganda: Detecting Spam Users in CollaborativeFiltering[C]//Proceedings of the12th International Conference on Intelligent User Interfaces. NewYork: ACM,2007:14-21.
[39] Mehta B. Unsupervised Shilling Detection for Collaborative Filtering[C]//Proceedings of the22ndNational Conference on Artificial intelligence. Menlo Park, CA: AAAI,2007:1402-1407.
[40] Bryan K, O’Mahony M, Cunningham P. Unsupervised Retrieval of Attack Profiles inCollaborative Recommender Systems[C]//Proceedings of the2008ACM Conference onRecommender Systems. New York: ACM,2008:155-162.
[41] Bhaumik R, Mobasher B, Burke R. A Clustering Approach to Unsupervised Attack Detection inCollaborative Recommender Systems[C]//Proceedings of the7th International Conference onData Mining. Washington: IEEE Computer Society,2011:181-187.
[42]李聪,骆志刚,石金龙.一种探测推荐系统托攻击的无监督算法[J].自动化学报,2011,37(2):160-167.
[43] Lee J S, Zhu D. Shilling Attack Detection-A New Approach for A Trustworthy RecommenderSystem[J]. JNFORMS Journal on Computing,2012,24(1):117-131.
[44] Chung C Y, Hsu P Y, Huang S H. βP: A Novel Approach to Filter out Malicious Rating Profilesfrom Recommender Systems[J]. Decision Support Systems,2013,55(1):314-325.
[45] Williams C A, Mobasher B, Burke R, et al. Detecting Profile Injection Attacks in CollaborativeFiltering: A Classification-Based Approach[C]//Proceedings of the8th Knowledge Discovery onthe Web International Conference on Advances in Web Mining and Web Usage Analysis. Berlin:Springer,2007:167-186.
[46] Williams C A, Mobasher B, Burke R. Defending Recommender Systems: Detection of ProfileInjection Attacks[J]. Service Oriented Computing and Applications,2007,1(3):157-170.
[47] He F, Wang X, Liu B. Attack Detection by Rough Set Theory in RecommendationSystem[C]//2010IEEE International Conference on Granular Computing. Washington: IEEEComputer Society,2010:692-695.
[48]伍之昂,庄毅,王有权,等.基于特征选择的推荐系统托攻击检测算法[J].电子学报,2012,40(8):1687-1693.
[49] Wu Z A, Gao J, Mao B, et al. Semi-SAD: Applying Semi-Supervised Learning to Shilling AttackDetection[C]//Proceedings of the5th ACM Conference on Recommender Systems. New York:ACM,2011:289-292.
[50] Wu Z A, Wu J J, Cao J, et al. HySAD: A Semi-Supervised Hybrid Shilling Attack Detector forTrustworthy Product Recommendation[C]//Proceedings of the18th ACM SIGKDD InternationalConference on Knowledge Discovery and Data Mining. New York: ACM,2012:985-993.
[51] Cooley J W, Tukey J W. An Algorithm for the Machine Calculation of Complex Fourier Series[J].Mathematics of Computation,1965,19(90):297-301.
[52] Stone H S, Orchard M T, Chang E C, et al. A Fast Direct Fourier-Based Algorithm for SubpixelRegistration of Images[J]. IEEE Transactions on Geoscience and Remote Sensing,2001,39(10):2235-2243.
[53] Srinivasa R B, Chatterji B N. An FFT-Based Technique for Translation, Rotation, andScale-Invariant Image Registration[J]. IEEE Transactions on Image Processing,1996,5(8):1266-1271.
[54] Stone H S, Wolpov R. Blind Cross-Spectral Image Registration Using Prefiltering andFourier-Based Translation Detection[J]. IEEE Transactions on Geosciences and Remote Sensing,2002,40(3):637-650.
[55]邹红星,周小波,李衍达.时频分析:回溯与前瞻[J].电子学报,2000,28(9):78-84.
[56] Huang N E, Shen Z, Long S R, et al. The Empirical Mode Decomposition and the HilbertSpectrum for Nonlinear and Non-Stationary Time Series Analysis[J]. Proceedings of the RoyalSociety of London A.1998,454(1971):903-995.
[57]杨志华,齐东旭,杨力华,等.基于经验模式分解的汉字字体识别方法[J].软件学报,2005,16(8):1438-1444.
[58]谭舜泉,黄继武,杨志华.基于Hilbert-Huang变换的JPEG2000隐写分析[J].计算机学报,2006,29(9):1702-1710.
[59]李淑芳,周卫东,蔡冬梅,等. EMD和SVM结合的脑电信号分类方法[J].生物医学工程学杂志,2011,28(5):891-894.
[60] Li H L, Kwong S, Yang L H, et al. Hilbert-Huang Transform for Analysis of Heart Rate Variabilityin Cardiac Health[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics,2011,8(6):1557-1567.
[61] Agrafioti F, Hatzinakos D, Anderson A K. ECG Pattern Analysis for Emotion Detection[J]. IEEETransactions on Affective Computing,2012,3(1):102-115.
[62] Hughes J M, Mao D, Rockmore D N, et al. Empirical Mode Decomposition Analysis for VisualStylometry[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(11):2147-2157.
[63] Salton G. Developments in Automatic Text Retrieval[J]. Science,1991,253(5023):974-980.
[64] Zhong N, Li Y F, Wu S T. Effective Pattern Discovery for Text Mining[J]. IEEE Transactions onKnowledge and Data Engineering,2012,24(1):30-44.
[65]吴飞,韩亚洪,庄越挺,等.图像-文本相关性挖掘的Web图像聚类方法[J].软件学报,2010,21(7):1561-1575.
[66]黄承慧,印鉴,侯昉.一种结合词项语义信息和TF-IDF方法的文本相似度量方法[J].计算机学报,2011,34(5):856-864.
[67] Yang Y M, Pedersen J O. A Comparative Study on Feature Selection in TextCategorization[C]//Proceedings of14th International Conference on Machine Learning. SanFrancisco, CA: Morgan Kaufmann,1997:412-420.
[68] Church K W, Hanks P. Word Association Norms, Mutual Information, and Lexicography[J].Computational Linguistics,1990,16(1):22-29.
[69]卢振泰,冯衍秋,冯前进,等.基于等效子午面与互信息量的医学图像配准[J].计算机学报,2009,32(8):1611-1617.
[70]吕庆文,陈武凡.基于互信息量的图像分割[J].计算机学报,2006,29(2):296-301.
[71] Shannon C E. A Mathematical Theory of Communication[J]. Bell System Technical Journal,1948,27(4):379-423.
[72] Shannon C E. A Mathematical Theory of Communication[J]. Bell System Technical Journal,1948,27(4):623-659.
[73]丁世飞,朱红,许新征,等.基于熵的模糊信息测试研究[J].计算机学报,2012,35(4):796-801.
[74]王国胤,张清华.不同知识粒度下粗糙集的不确定性研究[J].计算机学报,2008,31(9):1588-1598.
[75]谢宏,程浩忠,牛东晓.基于信息熵的粗糙集连续属性离散化算法[J].计算机学报,2005,28(9):1570-1574.
[76]范丽敏,冯登国,陈华.基于熵的随机性检测相关性研究[J].软件学报,2009,20(7):1967-1976.
[77]张文泉,张世英,江立勤.基于熵的决策评价模型及应用[J].系统工程学报,1995,10(3):69-74.
[78] Burges C J C. A Tutorial on Support Vector Machines for Pattern Recognition[J]. Data Mining andKnowledge Discovery,1998,2(2):121-167.
[79] Sengur A. Support Vector Machine Ensembles for Intelligent Diagnosis of Valvular HeartDisease[J]. Journal of Medical Systems,2012,36(4):2649-2655.
[80]曾志强,高济.基于向量集约简的精简支持向量机[J].软件学报,2007,18(11):2719-2727.
[81]饶鲜,董春曦,杨绍全.基于支持向量机的入侵检测系统[J].软件学报,2003,14(4):798-803.
[82]王瑞平,陈杰,山世光,等.基于支持向量机的人脸检测训练集增强[J].软件学报,2008,19(11):2921-2931.
[83] Golmohammadi H, Dashtbozorgi Z, Jr W E A. Quantitative Structure-Activity RelationshipPrediction of Blood-to-brain Partitioning Behavior Using Support Vector Machine[J]. EuropeanJournal of Pharmaceutical Sciences,2012,47(2):421-429.
[84] Wang B H, Huang H J, Wang X L. A Support Vector Machine Based MSM Model for FinancialShort-term Volatility Forecasting[J]. Neural Computing and Applications,2013,22(1):21-28.
[85] Christmann A, Hable R. Consistency of Support Vector Machines Using Additive Kernels forAdditive Models[J]. Computational Statistics&Data Analysis,2012,56(4):854-873.
[86] Wang Y R, Yu C Y, Chan H H. Predicting Construction Cost and Schedule Success Using ArtificialNeural Networks Ensemble and Support Vector Machines Classification Models[J]. InternationalJournal of Project Management,2012,30(4):470-478.
[87] Manupati V K, Anand R, Thakkar J J, et al. Adaptive Production Control System for a FlexibleManufacturing Cell Using Support Vector Machine-Based Approach[J]. The International Journalof Advanced Manufacturing Technology,2013,67(1-4):969-981.
[88] Alsulaiman F A, Sakr N, Valdes J J, et al. Identity Verification Based on Handwritten Signatureswith Haptic Information Using Genetic Programming[J]. ACM Transactions on MultimediaComputing, Communications and Applications,2013,9(2):1-21.
[89] Diosan L, Rogozan A, Pecuchet J P. Improving Classification Performance of Support VectorMachine by Genetically Optimising Kernel Shape and Hyper-Parameters[J]. Applied Intelligence,2012,36(2):280-294.
[90] Amini H, Gholami R, Monjezi M, et al. Evaluation of Flyrock Phenomenon Due to BlastingOperation by Support Vector Machine[J]. Neural Computing and Applications,2012,21(8):2077-2085.
[91] Dietterich T G. Machine Learning Research: Four Current Directions[J]. AI Magazine,1997,18(4):97-136.
[92] Verma B, Rahman A. Cluster-Oriented Ensemble Classifier: Impact of MulticlusterCharacterization on Ensemble Classifier Learning[J]. IEEE Transactions on Knowledge and DataEngineering,2012,24(4):605-618.
[93] Palit I, Reddy C K. Scalable and Parallel Boosting with MapReduce[J]. IEEE Transactions onKnowledge and Data Engineering,2012,24(10):1904-1916.
[94] Kocaguneli E, Menzies T, Keung J W. On the Value of Ensemble Effort Estimation[J]. IEEETransactions on Software Engineering,2012,38(6):1403-1416.
[95] Valentini G. True Path Rule Hierarchical Ensembles for Genome-Wide Gene FunctionPrediction[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics,2011,8(3):832-847.
[96] Oh S, Lee M S, Zhang B T. Ensemble Learning with Active Example Selection for ImbalancedBiomedical Data Classification[J]. IEEE/ACM Transactions on Computational Biology andBioinformatics,2011,8(2):316-325.
[97] Masud M M, Al-Khateeb T M, Hamlen K W, et al. Cloud-Based Malware Detection for EvolvingData Streams[J]. ACM Transactions on Management Information Systems,2011,2(3):1-27.
[98] Zhang X S, Shrestha B, Yoon S, et al. An Ensemble Architecture for Learning ComplexProblem-Solving Techniques from Demonstration[J]. ACM Transactions on Intelligent Systemsand Technology,2012,3(4):1-38.
[99] Kuncheva L I. A Bound on Kappa-Error Diagrams for Analysis of Classifier Ensembles[J]. IEEETransactions on Knowledge and Data Engineering,2013,25(3):494-501.
[100]Minku L L, Yao X. DDD: A New Ensemble Approach for Dealing with Concept Drift[J]. IEEETransactions on Knowledge and Data Engineering,2012,24(4):619-633.
[101]Wang S, Yao X. Relationships Between Diversity of Classification Ensembles and Single-ClassPerformance Measures[J]. IEEE Transactions on Knowledge and Data Engineering,2013,25(1):206-219.
[102]Breiman L. Bagging Predictors[J]. Machine Learning,1996,24(2):123-140.
[103]Kotsiantis S B. An Incremental Ensemble of Classifiers[J]. Artificial Intelligence Review,2011,36(4):249-266.
[104]Ho T K. The Random Subspace Method for Constructing Decision Forests[J]. IEEE Transactionson Pattern Analysis and Machine Intelligence,1998,20(8):832-844.
[105]Johnson R W. An Introduction to the Bootstrap[J]. Teaching Statistics,2001,23(2):49-54.
[106]Boos D D. Introduction to the Bootstrap World[J]. Statistical Science,2003,18(2):168-174.
[107]Panov P, Dzeroski S. Combining Bagging and Random Subspaces to Create BetterEnsembles[C]//Proceedings of the7th International Conference on Intelligent Data Analysis.Berlin: Springer,2007:118-129.
[108]张宏莉,鲁刚.分类不平衡协议流的机器学习算法评估与比较[J].软件学报,2012,23(6):1500-1516.
[109]Ehm W. Binomial Approximation to The Poisson Binomial Distribution[J]. Statistics andProbability Letters,1991,11(1):7-16.
[110]Majsnerowska M. A Note on Poisson Approximation by w-Functions[J]. ApplicationesMathematicae,1998,25(3):387-392.
[111]王守觉.仿生模式识别(拓扑模式识别)-一种模式识别新模型的理论与应用[J].电子学报,2002,30(10):1417-1420.
[112]王守觉,曲延锋,李卫军,等.基于仿生模式识别与传统模式识别的人脸识别效果比较研究[J].电子学报,2004,32(7):1057-1061.
[113]王守觉,孙华,柳培忠,等.基于仿生形象思维方法的图像检索算法[J].电子学报,2010,38(5):993-997.
[114]柳培忠,王守觉.利用多维空间同源连续性的图像检索[J].应用科学学报,2011,29(2):154-158.
[115]王守觉,潘晓霞,徐春燕,等.一种基于高维空间覆盖动态搜索方法的非特定人连续数字语音识别的研究[J].电子学报,2005,33(10),1790-1793.
[116]黄琦,魏建明,刘海涛.基于仿生模式识别的地面声目标识别方法[J].电子测量与仪器学报,2007,21(2):62-65.
[117]王守觉,曲延锋,李卫军,等.基于仿生模式识别与传统模式识别的人脸识别效果比较研究[J].电子学报,2004,32(7):1057-1061.
[118]安冬,王库,王守觉.高维空间点覆盖方法在物种计算机自动分类中的应用[J].电子学报,2006,34(2):277-281.
[119]翟亚锋,苏谦,邬文锦,等.基于仿生模式识别和近红外光谱的转基因小麦快速鉴别方法[J].光谱学与光谱分析,2010,30(4):924-928.
[120]王宪保,周德龙,王守觉.基于仿生模式识别的构造型神经网络分类方法[J].计算机学报,2007,30(12):2109-2114.
[121]Mehta B, Nejdl W. Attack Resistant Collaborative Filtering[C]//Proceedings of the31st AnnualInternational ACM SIGIR Conference on Research and Development in Information Retrieval.New York: ACM,2008:75-82.
[122]Chang C C, Lin C J. LIBSVM: A Library for Support Vector Machines[J]. ACM Transactions onIntelligent Systems and Technology,2011,2(3):1-27.
[123]Ambert K H, Cohen A M. K-Information Gain Scaled Nearest Neighbors: A Novel Approach toClassifying Protein-Protein Interaction-Related Documents[J]. IEEE/ACM Transactions onComputational Biology and Bioinformatics,2012,9(1):305-310.
[124]Fawcett T. An Introduction to ROC Analysis[J]. Pattern Recognition Letters,2006,27(8):861-874.
[125]Hand D J, Till R J. A Simple Generalisation of the Area Under the ROC Curve for Multiple ClassClassification Problems[J]. Machine Learning,2001,45(2):171-186.
[126]苏金树,张博锋,徐昕.基于机器学习的文本分类技术研究进展[J].软件学报,2006,17(9):1848-1859.
[127]Wasikowski M, Chen X W. Combating the Small Sample Class Imbalance Problem Using FeatureSelection[J]. IEEE Transactions on Knowledge and Data Engineering,2010,22(10):1388-1400.