Pattern Recognition and Machine Learning : Proceedings of the Japan--U.S. Seminar on the Learning Process in Control Systems, held in Nagoya, Japan August 18-20, 1970

This book contains the Proceedings of the US-Japan Seminar on Learning Process in Control Systems. The seminar, held in Nagoya, Japan, from August 18 to 20, 1970, was sponsored by the US-Japan Cooperative Science Program, jointly supported by the National Science Foundation and the Japan Society for the Promotion of Science. The full texts of all the presented papers except two t are included. The papers cover a great variety of topics related to learning processes and systems, ranging from pattern recognition to systems identification, from learning control to biological modelling. In order to reflect the actual content of the book, the present title was selected. All the twenty-eight papers are roughly divided into two parts--Pattern Recognition and System Identification and Learning Process and Learning Control. It is sometimes quite obvious that some papers can be classified into either part. The choice in these cases was strictly the editor's in order to keep a certain balance between the two parts. During the past decade there has been a considerable growth of interest in problems of pattern recognition and machine learn­ ing. In designing an optimal pattern recognition or control system, if all the a priori information about the process under study is known and can be described deterministically, the optimal system is usually designed by deterministic optimization techniques. Front Matter....Pages i-ix Some Studies on Pattern Recognition with Nonsupervised Learning Procedures....Pages 1-17 Linear and Nonlinear Stochastic Approximation Algorithms for Learning Systems....Pages 18-28 Multi-Category Pattern Classification Using a Nonsupervised Learning Algorithm....Pages 29-41 A Mixed-Type Non-Parametric Learning Machine without a Teacher....Pages 42-55 Recognition System for Handwritten Letters Simulating Visual Nervous System....Pages 56-69 Sequential Identification by Means of Gradient Learning Algorithms....Pages 70-78 Stochastic Approximation Algorithms for System Identification Using Normal Operating Data....Pages 79-86 On Utilization of Structural Information to Improve Identification Accuracy....Pages 87-96 An Inconsistency between the Rate and the Accuracy of the Learning Method for System Identification and its Tracking Characteristics....Pages 97-110 Weighting Function Estimation in Distributed-Parameter Systems....Pages 111-120 System Identifications by a Nonlinear Filter....Pages 121-137 A Linear Filter for Discrete Systems with Correlated Measurement Noise....Pages 138-149 Stochastic Learning by Means of Controlled Stochastic Processes....Pages 150-159 Learning Processes in a Random Machine....Pages 160-171 Learning Process in a Model of Associative Memory....Pages 172-186 Adaptive Optimization in Learning Control....Pages 187-194 Learning Control of Multimodal Systems by Fuzzy Automata....Pages 195-203 On a Class of Performance-Adaptive Self-Organizing Control Systems....Pages 204-220 A Control System Improving its Control Dynamics by Learning....Pages 221-229 Self-Learning Method for Time-Optimal Control....Pages 230-242 Learning Control via Associative Retrieval and Inference....Pages 243-251 Statistical Decision Method in Learning Control Systems....Pages 252-262 A Continuous-Valued Learning Controller for the global optimization of Stochastic Control Systems....Pages 263-276 On Variable-Structure Stochastic Automata....Pages 277-287 A Critical Review of Learning Control Research....Pages 288-296 Heuristics and Learning Control (Introduction to Intelligent Control)....Pages 297-309 Adaptive Model Control Applied to Real-Time Blood-Pressure Regulation....Pages 310-324 Real-Time Display System of Response Characteristics of Manual Control Systems....Pages 325-335 Back Matter....Pages 337-343 I: Pattern Recognition and System Identification.- Some Studies on Pattern Recognition with Nonsupervised Learning Procedures.- Linear and Nonlinear Stochastic Approximation Algorithms for Learning Systems.- Multi-Category Pattern Classification Using a Nonsupervised Learning Algorithm.- A Mixed-Type Non-Parametric Learning Machine Without a Teacher.- Recognition System for Handwritten Letters Simulating Visual Nervous System.- Sequential Identification by Means of Gradient Learning Algorithms.- Stochastic Approximation Algorithms for System Identification Using Normal Operating Data.- On Util