A critique of the valiant model
作者： Wray Buntine
DOI:
关键词: Task (project management) 、 Case analysis 、 Econometrics 、 Common sense 、 Mathematical economics 、 Bayesian probability 、 Mathematics 、 Interpretation (logic) 、 Sample (statistics) 、 Process (engineering) 、 Current (mathematics)
摘要: This paper considers the Valiant framework as it is applied to task of learning logical concepts from random examples. It argued that current interpretation this model departs common sense and practical experience in a number ways: does not allow sample dependent bounds, uses worst case rather than an average analysis, accommodate preferences about hypotheses. claimed result, can produce overlyconservative estimates confidence fail induction process often implemented. A Bayesian approach developed, based on notion disagreement between consistent seems overcome indicated problems.
ijcai.org 参考文章(22)
Jonathan Amsterdam, Some philosophical problems with formal learning theory national conference on artificial intelligence. pp. 580- 584 ,(1988)
JONATHAN AMSTERDAM, Extending the Valiant Learning Model international conference on machine learning. pp. 381- 394 ,(1988) , 10.1016/B978-0-934613-64-4.50044-X
DOUGLAS H. FISHER, JEFFREY C. SCHLIMMER, Concept simplification and prediction accuracy international conference on machine learning. pp. 22- 28 ,(1988) , 10.1016/B978-0-934613-64-4.50007-4
E Mark Gold, Language identification in the limit Information & Computation. ,vol. 10, pp. 447- 474 ,(1967) , 10.1016/S0019-9958(67)91165-5
James O Berger, Statistical Decision Theory and Bayesian Analysis ,(1993)
David Haussler, Quantifying inductive bias: AI learning algorithms and Valiant's learning framework Artificial Intelligence. ,vol. 36, pp. 177- 221 ,(1988) , 10.1016/0004-3702(88)90002-1
L. G. Valiant, A theory of the learnable symposium on the theory of computing. ,vol. 27, pp. 1134- 1142 ,(1984) , 10.1145/800057.808710
Anselm Blumer, Andrzej Ehrenfeucht, David Haussler, Manfred K. Warmuth, Occam's razor Information Processing Letters. ,vol. 24, pp. 377- 380 ,(1987) , 10.1016/0020-0190(87)90114-1
Dana Angluin, Philip Laird, Learning From Noisy Examples Machine Learning. ,vol. 2, pp. 343- 370 ,(1988) , 10.1023/A:1022873112823
Nick Littlestone, Learning Quickly When Irrelevant Attributes Abound: A New Linear-Threshold Algorithm Machine Learning. ,vol. 2, pp. 285- 318 ,(1988) , 10.1023/A:1022869011914