Machine Learning: ECML 2003: 14th European Conference on Machine Learning, Cavtat-Dubrovnik, Croatia, September 22-26, 2003, Proceedings (Lecture Notes in Computer Science, 2837)

The Proceedings Of Ecml/pkdd2003 Are Published In Two Volumes: The P- Ceedings Of The 14th European Conference On Machine Learning (lnai 2837) And The Proceedings Of The 7th European Conference On Principles And Practice Of Knowledge Discovery In Databases (lnai 2838). The Two Conferences Were Held On September 22–26, 2003 In Cavtat, A Small Tourist Town In The Vicinity Of Dubrovnik, Croatia. As Machine Learning And Knowledge Discovery Are Two Highly Related ?elds, Theco-locationofbothconferencesisbene?cialforbothresearchcommunities.in Cavtat, Ecml And Pkdd Were Co-located For The Third Time In A Row, Following The Successful Co-location Of The Two European Conferences In Freiburg (2001) And Helsinki (2002). The Co-location Of Ecml2003 And Pkdd2003 Resulted In A Joint Program For The Two Conferences, Including Paper Presentations, Invited Talks, Tutorials, And Workshops. Out Of 332 Submitted Papers, 40 Were Accepted For Publication In The Ecml2003proceedings,and40wereacceptedforpublicationinthepkdd2003 Proceedings. All The Submitted Papers Were Reviewed By Three Referees. In Ad- Tion To Submitted Papers, The Conference Program Consisted Of Four Invited Talks, Four Tutorials, Seven Workshops, Two Tutorials Combined With A Workshop, And A Discovery Challenge. Invited Papers -- From Knowledge-based To Skill-based Systems: Sailing As A Machine Learning Challenge -- Two-eyed Algorithms And Problems -- Next Generation Data Mining Tools: Power Laws And Self-similarity For Graphs, Streams And Traditional Data -- Taking Causality Seriously: Propensity Score Methodology Applied To Estimate The Effects Of Marketing Interventions -- Contributed Papers -- Support Vector Machines With Example Dependent Costs -- Abalearn: A Risk-sensitive Approach To Self-play Learning In Abalone -- Life Cycle Modeling Of News Events Using Aging Theory -- Unambiguous Automata Inference By Means Of State-merging Methods -- Could Active Perception Aid Navigation Of Partially Observable Grid Worlds? -- Combined Optimization Of Feature Selection And Algorithm Parameters In Machine Learning Of Language -- Iteratively Extending Time Horizon Reinforcement Learning -- Volume Under The Roc Surface For Multi-class Problems --^ Improving The Auc Of Probabilistic Estimation Trees -- Scaled Cgem: A Fast Accelerated Em -- Pairwise Preference Learning And Ranking -- A New Way To Introduce Knowledge Into Reinforcement Learning -- Improvement Of The State Merging Rule On Noisy Data In Probabilistic Grammatical Inference -- Collective Intelligence With Sequences Of Actions -- Rademacher Penalization Over Decision Tree Prunings -- Learning Rules To Improve A Machine Translation System -- Optimising Performance Of Competing Search Engines In Heterogeneous Web Environments -- Robust K-dnf Learning Via Inductive Belief Merging -- Logistic Model Trees -- Color Image Segmentation: Kernel Do The Feature Space -- Evaluation Of Topographic Clustering And Its Kernelization -- A New Pairwise Ensemble Approach For Text Classification -- Self-evaluated Learning Agent In Multiple State Games -- Classification Approach Towards Ranking And Sorting Problems --^ Using Mdp Characteristics To Guide Exploration In Reinforcement Learning -- Experiments With Cost-sensitive Feature Evaluation -- A Markov Network Based Factorized Distribution Algorithm For Optimization -- On Boosting Improvement: Error Reduction And Convergence Speed-up -- Improving Svm Text Classification Performance Through Threshold Adjustment -- Backoff Parameter Estimation For The Dop Model -- Improving Numerical Prediction With Qualitative Constraints -- A Generative Model For Semantic Role Labeling -- Optimizing Local Probability Models For Statistical Parsing -- Extended Replicator Dynamics As A Key To Reinforcement Learning In Multi-agent Systems -- Visualizations For Assessing Convergence And Mixing Of Mcmc -- A Decomposition Of Classes Via Clustering To Explain And Improve Naive Bayes -- Improving Rocchio With Weakly Supervised Clustering -- A Two-level Learning Method For Generalized Multi-instance Problems -- Clustering In Knowledge Embedded Space --^ Ensembles Of Multi-instance Learners. Nada Lavrač ... [et Al.] (eds.). Includes Bibliographical References And Index. Front Matter....Pages - From Knowledge-Based to Skill-Based Systems: Sailing as a Machine Learning Challenge....Pages 1-8 Two-Eyed Algorithms and Problems....Pages 9-9 Next Generation Data Mining Tools: Power Laws and Self-similarity for Graphs, Streams and Traditional Data....Pages 10-15 Taking Causality Seriously: Propensity Score Methodology Applied to Estimate the Effects of Marketing Interventions....Pages 16-22 Support Vector Machines with Example Dependent Costs....Pages 23-34 Abalearn: A Risk-Sensitive Approach to Self-play Learning in Abalone....Pages 35-46 Life Cycle Modeling of News Events Using Aging Theory....Pages 47-59 Unambiguous Automata Inference by Means of State-Merging Methods....Pages 60-71 Could Active Perception Aid Navigation of Partially Observable Grid Worlds?....Pages 72-83 Combined Optimization of Feature Selection and Algorithm Parameters in Machine Learning of Language....Pages 84-95 Iteratively Extending Time Horizon Reinforcement Learning....Pages 96-107 Volume under the ROC Surface for Multi-class Problems....Pages 108-120 Improving the AUC of Probabilistic Estimation Trees....Pages 121-132 Scaled CGEM: A Fast Accelerated EM....Pages 133-144 Pairwise Preference Learning and Ranking....Pages 145-156 A New Way to Introduce Knowledge into Reinforcement Learning....Pages 157-168 Improvement of the State Merging Rule on Noisy Data in Probabilistic Grammatical Inference....Pages 169-180 COllective INtelligence with Sequences of Actions....Pages 181-192 Rademacher Penalization over Decision Tree Prunings....Pages 193-204 Learning Rules to Improve a Machine Translation System....Pages 205-216 Optimising Performance of Competing Search Engines in Heterogeneous Web Environments....Pages 217-228 Robust k -DNF Learning via Inductive Belief Merging....Pages 229-240 Logistic Model Trees....Pages 241-252 Color Image Segmentation: Kernel Do the Feature Space....Pages 253-264 Evaluation of Topographic Clustering and Its Kernelization....Pages 265-276 A New Pairwise Ensemble Approach for Text Classification....Pages 277-288 Self-evaluated Learning Agent in Multiple State Games....Pages 289-300 Classification Approach towards Ranking and Sorting Problems....Pages 301-312 Using MDP Characteristics to Guide Exploration in Reinforcement Learning....Pages 313-324 Experiments with Cost-Sensitive Feature Evaluation....Pages 325-336 A Markov Network Based Factorized Distribution Algorithm for Optimization....Pages 337-348 On Boosting Improvement: Error Reduction and Convergence Speed-Up....Pages 349-360 Improving SVM Text Classification Performance through Threshold Adjustment....Pages 361-372 Backoff Parameter Estimation for the DOP Model....Pages 373-384 Improving Numerical Prediction with Qualitative Constraints....Pages 385-396 A Generative Model for Semantic Role Labeling....Pages 397-408 Optimizing Local Probability Models for Statistical Parsing....Pages 409-420 Extended Replicator Dynamics as a Key to Reinforcement Learning in Multi-agent Systems....Pages 421-431 Visualizations for Assessing Convergence and Mixing of MCMC....Pages 432-443 A Decomposition of Classes via Clustering to Explain and Improve Naive Bayes....Pages 444-455 Improving Rocchio with Weakly Supervised Clustering....Pages 456-467 A Two-Level Learning Method for Generalized Multi-instance Problems....Pages 468-479 Clustering in Knowledge Embedded Space....Pages 480-491 Ensembles of Multi-instance Learners....Pages 492-502 Back Matter....Pages - The proceedings of ECML/PKDD2003 are published in two volumes: the P- ceedings of the 14th European Conference on Machine Learning (LNAI 2837) and the Proceedings of the 7th European Conference on Principles and Practice of Knowledge Discovery in Databases (LNAI 2838). The two conferences were held on September 22{u2013}26, 2003 in Cavtat, a small tourist town in the vicinity of Dubrovnik, Croatia. As machine learning and knowledge discovery are two highly related ?elds, theco-locationofbothconferencesisbene?cialforbothresearchcommunities.In Cavtat, ECML and PKDD were co-located for the third time in a row, following the successful co-location of the two European conferences in Freiburg (2001) and Helsinki (2002). The co-location of ECML2003 and PKDD2003 resulted in a joint program for the two conferences, including paper presentations, invited talks, tutorials, and workshops. Out of 332 submitted papers, 40 were accepted for publication in the ECML2003proceedings,and40wereacceptedforpublicationinthePKDD2003 proceedings. All the submitted papers were reviewed by three referees. In ad- tion to submitted papers, the conference program consisted of four invited talks, four tutorials, seven workshops, two tutorials combined with a workshop, and a discovery challenge This book constitutes the refereed proceedings of the 14th European Conference on Machine Learning, ECML 2003, held in Cavtat-Dubrovnik, Croatia in September 2003 in conjunction with PKDD 2003. The 40 revised full papers presented together with 4 invited contributions were carefully reviewed and, together with another 40 ones for PKDD 2003, selected from a total of 332 submissions. The papers address all current issues in machine learning including support vector machine, inductive inference, feature selection algorithms, reinforcement learning, preference learning, probabilistic grammatical inference, decision tree learning, clustering, classification, agent learning, Markov networks, boosting, statistical parsing, Bayesian learning, supervised learning, and multi-instance learning