Machine Learning: ECML 2007: 18th European Conference on Machine Learning, Warsaw, Poland, September 17-21, 2007, Proceedings (Lecture Notes in Computer Science (4701))

The two premier annual European conferences in the areas of machine learning and data mining have been collocated ever since the ?rst joint conference in Freiburg, 2001. The European Conference on Machine Learning (ECML) traces its origins to 1986, when the ?rst European Working Session on Learning was held in Orsay, France. The European Conference on Principles and Practice of KnowledgeDiscoveryinDatabases(PKDD) was?rstheldin1997inTrondheim, Norway. Over the years, the ECML/PKDD series has evolved into one of the largest and most selective international conferences in machine learning and data mining. In 2007, the seventh collocated ECML/PKDD took place during September 17–21 on the centralcampus of WarsawUniversityand in the nearby Staszic Palace of the Polish Academy of Sciences. The conference for the third time used a hierarchical reviewing process. We nominated 30 Area Chairs, each of them responsible for one sub-?eld or several closely related research topics. Suitable areas were selected on the basis of the submission statistics for ECML/PKDD 2006 and for last year’s International Conference on Machine Learning (ICML 2006) to ensure a proper load balance amongtheAreaChairs.AjointProgramCommittee(PC)wasnominatedforthe two conferences, consisting of some 300 renowned researchers, mostly proposed by the Area Chairs. This joint PC, the largest of the series to date, allowed us to exploit synergies and deal competently with topic overlaps between ECML and PKDD. ECML/PKDD 2007 received 592 abstract submissions. As in previous years, toassistthereviewersandtheAreaChairsintheir?nalrecommendationauthors had the opportunity to communicate their feedback after the reviewing phase. Front Matter....Pages - Learning, Information Extraction and the Web....Pages 1-1 Putting Things in Order: On the Fundamental Role of Ranking in Classification and Probability Estimation....Pages 2-3 Mining Queries....Pages 4-4 Adventures in Personalized Information Access....Pages 5-5 Statistical Debugging Using Latent Topic Models....Pages 6-17 Learning Balls of Strings with Correction Queries....Pages 18-29 Neighborhood-Based Local Sensitivity....Pages 30-41 Approximating Gaussian Processes with ${\cal H}^2$ -Matrices....Pages 42-53 Learning Metrics Between Tree Structured Data: Application to Image Recognition....Pages 54-66 Shrinkage Estimator for Bayesian Network Parameters....Pages 67-78 Level Learning Set: A Novel Classifier Based on Active Contour Models....Pages 79-90 Learning Partially Observable Markov Models from First Passage Times....Pages 91-103 Context Sensitive Paraphrasing with a Global Unsupervised Classifier....Pages 104-115 Dual Strategy Active Learning....Pages 116-127 Decision Tree Instability and Active Learning....Pages 128-139 Constraint Selection by Committee: An Ensemble Approach to Identifying Informative Constraints for Semi-supervised Clustering....Pages 140-151 The Cost of Learning Directed Cuts....Pages 152-163 Spectral Clustering and Embedding with Hidden Markov Models....Pages 164-175 Probabilistic Explanation Based Learning....Pages 176-187 Graph-Based Domain Mapping for Transfer Learning in General Games....Pages 188-200 Learning to Classify Documents with Only a Small Positive Training Set....Pages 201-213 Structure Learning of Probabilistic Relational Models from Incomplete Relational Data....Pages 214-225 Stability Based Sparse LSI/PCA: Incorporating Feature Selection in LSI and PCA....Pages 226-237 Bayesian Substructure Learning - Approximate Learning of Very Large Network Structures....Pages 238-249 Efficient Continuous-Time Reinforcement Learning with Adaptive State Graphs....Pages 250-261 Source Separation with Gaussian Process Models....Pages 262-273 Discriminative Sequence Labeling by Z-Score Optimization....Pages 274-285 Fast Optimization Methods for L1 Regularization: A Comparative Study and Two New Approaches....Pages 286-297 Bayesian Inference for Sparse Generalized Linear Models....Pages 298-309 Classifier Loss Under Metric Uncertainty....Pages 310-322 Additive Groves of Regression Trees....Pages 323-334 Efficient Computation of Recursive Principal Component Analysis for Structured Input....Pages 335-346 Hinge Rank Loss and the Area Under the ROC Curve....Pages 347-358 Clustering Trees with Instance Level Constraints....Pages 359-370 On Pairwise Naive Bayes Classifiers....Pages 371-381 Separating Precision and Mean in Dirichlet-Enhanced High-Order Markov Models....Pages 382-393 Safe Q-Learning on Complete History Spaces....Pages 394-405 Random k -Labelsets: An Ensemble Method for Multilabel Classification....Pages 406-417 Seeing the Forest Through the Trees: Learning a Comprehensible Model from an Ensemble....Pages 418-429 Avoiding Boosting Overfitting by Removing Confusing Samples....Pages 430-441 Planning and Learning in Environments with Delayed Feedback....Pages 442-453 Analyzing Co-training Style Algorithms....Pages 454-465 Policy Gradient Critics....Pages 466-477 An Improved Model Selection Heuristic for AUC....Pages 478-489 Finding the Right Family: Parent and Child Selection for Averaged One-Dependence Estimators....Pages 490-501 Stepwise Induction of Multi-target Model Trees....Pages 502-509 Comparing Rule Measures for Predictive Association Rules....Pages 510-517 User Oriented Hierarchical Information Organization and Retrieval....Pages 518-526 Learning a Classifier with Very Few Examples: Analogy Based and Knowledge Based Generation of New Examples for Character Recognition....Pages 527-534 Weighted Kernel Regression for Predicting Changing Dependencies....Pages 535-542 Counter-Example Generation-Based One-Class Classification....Pages 543-550 Test-Cost Sensitive Classification Based on Conditioned Loss Functions....Pages 551-558 Probabilistic Models for Action-Based Chinese Dependency Parsing....Pages 559-566 Learning Directed Probabilistic Logical Models: Ordering-Search Versus Structure-Search....Pages 567-574 A Simple Lexicographic Ranker and Probability Estimator....Pages 575-582 On Minimizing the Position Error in Label Ranking....Pages 583-590 On Phase Transitions in Learning Sparse Networks....Pages 591-599 Semi-supervised Collaborative Text Classification....Pages 600-607 Learning from Relevant Tasks Only....Pages 608-615 An Unsupervised Learning Algorithm for Rank Aggregation....Pages 616-623 Ensembles of Multi-Objective Decision Trees....Pages 624-631 Kernel-Based Grouping of Histogram Data....Pages 632-639 Active Class Selection....Pages 640-647 Sequence Labeling with Reinforcement Learning and Ranking Algorithms....Pages 648-657 Efficient Pairwise Classification....Pages 658-665 Scale-Space Based Weak Regressors for Boosting....Pages 666-673 K -Means with Large and Noisy Constraint Sets....Pages 674-682 Towards ‘Interactive’ Active Learning in Multi-view Feature Sets for Information Extraction....Pages 683-690 Principal Component Analysis for Large Scale Problems with Lots of Missing Values....Pages 691-698 Transfer Learning in Reinforcement Learning Problems Through Partial Policy Recycling....Pages 699-707 Class Noise Mitigation Through Instance Weighting....Pages 708-715 Optimizing Feature Sets for Structured Data....Pages 716-723 Roulette Sampling for Cost-Sensitive Learning....Pages 724-731 Modeling Highway Traffic Volumes....Pages 732-739 Undercomplete Blind Subspace Deconvolution Via Linear Prediction....Pages 740-747 Learning an Outlier-Robust Kalman Filter....Pages 748-756 Imitation Learning Using Graphical Models....Pages 757-764 Nondeterministic Discretization of Weights Improves Accuracy of Neural Networks....Pages 765-772 Semi-definite Manifold Alignment....Pages 773-781 General Solution for Supervised Graph Embedding....Pages 782-789 Multi-objective Genetic Programming for Multiple Instance Learning....Pages 790-797 Exploiting Term, Predicate, and Feature Taxonomies in Propositionalization and Propositional Rule Learning....Pages 798-805 Back Matter....Pages - The two premier annual European conferences in the areas of machine learning and data mining have been collocated ever since the?rst joint conference in Freiburg, 2001. The European Conference on Machine Learning (ECML) traces its origins to 1986, when the?rst European Working Session on Learning was held in Orsay, France. The European Conference on Principles and Practice of KnowledgeDiscoveryinDatabases(PKDD) was?rstheldin1997inTrondheim, Norway. Over the years, the ECML/PKDD series has evolved into one of the largest and most selective international conferences in machine learning and data mining. In 2007, the seventh collocated ECML/PKDD took place during September 17-21 on the centralcampus of WarsawUniversityand in the nearby Staszic Palace of the Polish Academy of Sciences. The conference for the third time used a hierarchical reviewing process. We nominated 30 Area Chairs, each of them responsible for one sub-?eld or several closely related research topics. Suitable areas were selected on the basis of the submission statistics for ECML/PKDD 2006 and for last year's International Conference on Machine Learning (ICML 2006) to ensure a proper load balance amongtheAreaChairs. AjointProgramCommittee(PC)wasnominatedforthe two conferences, consisting of some 300 renowned researchers, mostly proposed by the Area Chairs. This joint PC, the largest of the series to date, allowed us to exploit synergies and deal competently with topic overlaps between ECML and PKDD. ECML/PKDD 2007 received 592 abstract submissions. As in previous years, toassistthereviewersandtheAreaChairsintheir?nalrecommendationauthors had the opportunity to communicate their feedback after the reviewing phase Joost N. Kok ... [et Al.] (eds.). Includes Bibliographical References And Index. Also Issued Online.