This volume is a selection of papers presented at the Fourth International Workshop on Artificial Intelligence and Statistics held in January 1993. These biennial workshops have succeeded in bringing together researchers from Artificial Intelligence and from Statistics to discuss problems of mutual interest. The exchange has broadened research in both fields and has strongly encour aged interdisciplinary work. The theme ofthe 1993 AI and Statistics workshop was: "Selecting Models from Data". The papers in this volume attest to the diversity of approaches to model selection and to the ubiquity of the problem. Both statistics and artificial intelligence have independently developed approaches to model selection and the corresponding algorithms to implement them. But as these papers make clear, there is a high degree of overlap between the different approaches. In particular, there is agreement that the fundamental problem is the avoidence of "overfitting"-Le., where a model fits the given data very closely, but is a poor predictor for new data; in other words, the model has partly fitted the "noise" in the original data.

Klappentext

This volume presents a selection of papers from the Fourth International Workshop on Artificial Intelligence and Statistics. This biennial workshop brings together researchers from both fields to discuss problems of mutual interest and to compare approaches to their solution. The fourth workshop focused on the topic of selecting models from data. As the papers in this volume attest, the empirical approaches from the two separate fields have much in common yet still depart enough from one another to stimulate active interdisciplinary work. The papers cover a wide spectrum of problems in empirical modelling including model selection in general, graphical models, causal models, regression and other statistical models, and general algorithms and software tools. This timely volume will benefit all researchers with an active interest in model selection, empirical model building, or more generally the interaction between Statistics and Artificial Intelligence.

Inhalt
I Overviews: Model Selection.- 1 Statistical strategy: step 1.- 2 Rational Learning: Finding a Balance Between Utility and Efficiency.- 3 A new criterion for selecting models from partially observed data.- 4 Small-sample and large-sample statistical model selection criteria.- 5 On the choice of penalty term in generalized FPE criterion.- 6 Cross-Validation, Stacking and Bi-Level Stacking: Meta-Methods for Classification Learning.- 7 Probabilistic approach to model selection: comparison with unstructured data set.- 8 Detecting and Explaining Dependencies in Execution Traces.- 9 A method for the dynamic selection of models under time constraints.- II Graphical Models.- 10 Strategies for Graphical Model Selection.- 11 Conditional dependence in probabilistic networks.- 12 Reuse and sharing of graphical belief network components.- 13 Bayesian Graphical Models for Predicting Errors in Databases.- 14 Model Selection for Diagnosis and Treatment Using Temporal Influence Diagrams.- 15 Diagnostic systems by model selection: a case study.- 16 A Survey of Sampling Methods for Inference on Directed Graphs.- 17 Minimizing decision table sizes in influence diagrams: dimension shrinking.- 18 Models from Data for Various Types of Reasoning.- III Causal Models.- 19 Causal inference in artificial intelligence.- 20 Inferring causal structure among unmeasured variables.- 21 When can association graphs admit a causal interpretation?.- 22 Inference, Intervention, and Prediction.- 23 Attitude Formation Models: Insights from TETRAD.- 24 Discovering Probabilistic Causal Relationships: A Comparison Between Two Methods.- 25 Path Analysis Models of an Autonomous Agent in a Complex Environment.- IV Particular Models.- 26 A Parallel Constructor of Markov Networks.- 27 Capturing observations in a nonstationary hidden Markov model.- 28 Extrapolating Definite Integral Information.- 29 The Software Reliability Consultant.- 30 Statistical Reasoning to Enhance User Modelling in Consulting Systems.- 31 Selecting a frailty model for longitudinal breast cancer data.- 32 Optimal design of reflective sensors using probabilistic analysis.- V Similarity-Based Models.- 33 Learning to Catch: Applying Nearest Neighbor Algorithms to Dynamic Control Tasks.- 34 Dynamic Recursive Model Class Selection for Classifier Construction.- 35 Minimizing the expected costs of classifying patterns by sequential costly inspections.- 36 Combining a knowledge-based system and a clustering method for a construction of models in ill-structured domains.- 37 Clustering of Symbolically Described Events for Prediction of Numeric Attributes.- 38 Symbolic Classifiers: Conditions to Have Good Accuracy Performance.- VI Regression and Other Statistical Models.- 39 Statistical and neural network techniques for nonparametric regression.- 40 Multicollinearity: A tale of two nonparametric regressions.- 41 Choice of Order in Regression Strategy.- 42 Modelling response models in software.- 43 Principal components and model selection.- VII Algorithms and Tools.- 44 Algorithmic speedups in growing classification trees by using an additive split criterion.- 45 Markov Chain Monte Carlo Methods for Hierarchical Bayesian Expert Systems.- 46 Simulated annealing in the construction of near-optimal decision trees.- 47 SA/GA: Survival of the Fittest in Alaska.- 48 A Tool for Model Generation and Knowledge Acquisition.- 49 Using knowledge-assisted discriminant analysis to generate new comparative terms.