Machine Learning: Principles and TechniquesRichard Forsyth Presents results of research into computer systems that can improve their own performance. For undergraduates, graduates, and professionals intending to write or use such systems. The various perspectives of over a dozen contributors are abstracted into the unifying principle: generate + test, which makes possible a provisional taxonomy of machine learning algorithms. The sections cover a background to induction, biologically inspired systems, automated discovery, and long-term perspectives. The paper edition ($29.95) was not seen. Annotation copyrighted by Book News, Inc., Portland, OR |
Contents
The logic of induction | 3 |
Machine induction as a form of knowledge acquisition | 23 |
the users perspective | 39 |
Copyright | |
14 other sections not shown
Common terms and phrases
acquisition AM's applied approach Artificial Intelligence attributes BACON basic causal Classical Classicists cluster quality CLUSTER/2 CMIT cognitive architecture cognitive science concepts Connectionism Connectionist models context creativity criteria cycle of discovery CYRANO database decision rules defined described descriptions developed discovery programs discriminant function domain elements engineering EURISKO evaluation evolution strategy examples excited expert systems Fodor and Pylyshyn genetic algorithm girl loves John given GLAUBER heuristics hierarchy human implementation implementation theory inductive learning inductive systems inference input instance intrinsic John loves jth stage laws learning process Lenat linear linguistic machine learning mechanism memory mental representations MeSH mutation natural language neural neurons Node notion occur operations organization output pair pattern problem produce properties random relations seed events segments selection semantic Silico Silico sapiens Snell's law specific stress supplied structure symbols syntactic TALE-SPIN theory values variables