A Rules-Based Approach for Configuring Chains of Classifiers in Real-Time Stream Mining Systems

Networks of classifiers can offer improved accuracy and scalability over single classifiers by utilizing distributed processing resources and analytics. However, they also pose a unique combination of challenges. First, classifiers may be located across different sites that are willing to cooperate to provide services, but are unwilling to reveal proprietary information about their analytics, or are unable to exchange their analytics due to the high transmission overheads involved. Furthermore, processing of voluminous stream data across sites often requires load shedding approaches, which can lead to suboptimal classification performance. Finally, real stream mining systems often exhibit dynamic behavior and thus necessitate frequent reconfiguration of classifier elements to ensure acceptable end-to-end performance and delay under resource constraints. Under such informational constraints, resource constraints, and unpredictable dynamics, utilizing a single, fixed algorithm for reconfiguring classifiers can often lead to poor performance. In this paper, we propose a new optimization framework aimed at developing rules for choosing algorithms to reconfigure the classifier system under such conditions. We provide an adaptive, Markov model-based solution for learning the optimal rule when stream dynamics are initially unknown. Furthermore, we discuss how rules can be decomposed across multiple sites and propose a method for evolving new rules from a set of existing rules. Simulation results are presented for a speech classification system to highlight the advantages of using the rules-based framework to cope with stream dynamics.

[1]  J. Nash THE BARGAINING PROBLEM , 1950, Classics in Game Theory.

[2]  P. Burke The Output of a Queuing System , 1956 .

[3]  Carl M. Harris,et al.  Fundamentals of queueing theory , 1975 .

[4]  E. M. Rounds A combined nonparametric approach to feature selection and binary decision tree design , 1980, Pattern Recognit..

[5]  Carl M. Harris,et al.  Fundamentals of queueing theory (2nd ed.). , 1985 .

[6]  Dimitri P. Bertsekas,et al.  Data Networks , 1986 .

[7]  Alfred O. Hero,et al.  Simultaneous signal detection and classification under a false alarm constraint , 1990, International Conference on Acoustics, Speech, and Signal Processing.

[8]  Eric Horvitz,et al.  Time-Dependent Utility and Action Under Uncertainty , 1991, UAI.

[9]  C. O'Cinneide Entrywise perturbation theory and error analysis for Markov chains , 1993 .

[10]  Martin L. Puterman,et al.  Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[11]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1997 .

[12]  Robert G. Gallager,et al.  Discrete Stochastic Processes , 1995 .

[13]  Richard S. Sutton,et al.  Introduction to Reinforcement Learning , 1998 .

[14]  Robert E. Schapire,et al.  A Brief Introduction to Boosting , 1999, IJCAI.

[15]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[16]  Tsuhan Chen,et al.  I/P frame selection using classification based mode decision , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[17]  Jeffrey F. Naughton,et al.  Rate-based query optimization for streaming information sources , 2002, SIGMOD '02.

[18]  Shih-Fu Chang,et al.  Structure analysis of soccer video with hidden Markov models , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[19]  Nesime Tatbul OoS-Driven Load Shedding on Data Streams , 2002, EDBT PhD Workshop.

[20]  Vladimir Pavlovic,et al.  Bayesian networks as ensemble of classifiers , 2002, Object recognition supported by user interaction for service robots.

[21]  Dustin Boswell,et al.  Introduction to Support Vector Machines , 2002 .

[22]  Joydeep Ghosh,et al.  Privacy-preserving distributed clustering using generative models , 2003, Third IEEE International Conference on Data Mining.

[23]  Luhong Liang,et al.  A detector tree of boosted classifiers for real-time object detection and tracking , 2003, 2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698).

[24]  Rajeev Motwani,et al.  Chain: operator scheduling for memory minimization in data stream systems , 2003, SIGMOD '03.

[25]  Michael Stonebraker,et al.  Load Shedding in a Data Stream Manager , 2003, VLDB.

[26]  Ying Xing,et al.  Scalable Distributed Stream Processing , 2003, CIDR.

[27]  Joseph M. Hellerstein,et al.  Flux: an adaptive partitioning operator for continuous query systems , 2003, Proceedings 19th International Conference on Data Engineering (Cat. No.03CH37405).

[28]  Jennifer Widom,et al.  Adaptive filters for continuous queries over distributed data streams , 2003, SIGMOD '03.

[29]  Frank Eliassen,et al.  Supporting timeliness and accuracy in distributed real-time content-based video analysis , 2003, MULTIMEDIA '03.

[30]  Mohamed Medhat Gaber,et al.  Cost-Efficient Mining Techniques for Data Streams , 2004, ACSW.

[31]  Ying Xing,et al.  Dynamic load distribution in the Borealis stream processor , 2005, 21st International Conference on Data Engineering (ICDE'05).

[32]  Michael Stonebraker,et al.  Fault-tolerance in the Borealis distributed stream processing system , 2005, SIGMOD '05.

[33]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[34]  Gary Weiss,et al.  Improving classifier utility by altering the misclassification cost ratio , 2005, UBDM '05.

[35]  Manuela M. Veloso,et al.  Learning to Select Negotiation Strategies in Multi-agent Meeting Scheduling , 2005, EPIA.

[36]  Philip S. Yu,et al.  Loadstar: A Load Shedding Scheme for Classifying Data Streams , 2005, SDM.

[37]  Karsten Schwan,et al.  Distributed Stream Management using Utility-Driven Self-Adaptive Middleware , 2005, Second International Conference on Autonomic Computing (ICAC'05).

[38]  Deepak S. Turaga,et al.  Resource Management for Networked Classifiers in Distributed Stream Mining Systems , 2006, Sixth International Conference on Data Mining (ICDM'06).

[39]  Fan Ye,et al.  Multi-site cooperative data stream analysis , 2006, OPSR.

[40]  Stanley B. Zdonik,et al.  Dealing with Overload in Distributed Stream Processing Systems , 2006, 22nd International Conference on Data Engineering Workshops (ICDEW'06).

[41]  D. Turaga,et al.  Resource Management for Chained Binary Classifiers , 2006 .

[42]  Mihaela van der Schaar,et al.  Configuring Competing Classifier Chains in Distributed Stream Mining Systems , 2007, IEEE Journal of Selected Topics in Signal Processing.

[43]  Rong Yan,et al.  IBM multimedia analysis and retrieval system , 2008, CIVR '08.

[44]  Brian Foo,et al.  Distributed classifier chain optimization for real-time multimedia stream mining systems , 2008, Electronic Imaging.

[45]  John R. Smith,et al.  IBM Research TRECVID-2009 Video Retrieval System , 2009, TRECVID.