摘要: Humans appear to be able learn new concepts without needing programmed explicitly in any conventional sense. In this paper we regard learning as the phenomenon of knowledge acquisition absence explicit programming. We give a precise methodology for studying from computational viewpoint. It consists choosing an appropriate information gathering mechanism, protocol, and exploring class that can learnt using it reasonable (polynomial) number steps. find inherent algorithmic complexity appears set serious limits range so learnt. The results suggest concrete principles designing realistic systems.
acm.org
core.ac.uk
mpg.de
lsu.edu 
princeton.edu 
nott.ac.uk 
brandeis.edu 
cumincad.org 
sci-hub.se 参考文章(8)
Joel Spencer, Probabilistic Methods in Combinatorics ,(1974)
J. G. Carbonell, T. M. Mitchell, R. S. Michalski, Machine Learning: An Artificial Intelligence Approach Springer Publishing Company, Incorporated. ,(2013)
Avron Barr, Edward A Feigenbaum, Paul R Cohen, The Handbook of Artificial Intelligence ,(1981)
Stephen A. Cook, The complexity of theorem-proving procedures symposium on the theory of computing. pp. 151- 158 ,(1971) , 10.1145/800157.805047
Dana Angluin, Carl H. Smith, Inductive Inference: Theory and Methods ACM Computing Surveys. ,vol. 15, pp. 237- 269 ,(1983) , 10.1145/356914.356918
Deductive Learning [and Discussion] Philosophical Transactions of the Royal Society A. ,vol. 312, pp. 441- 446 ,(1984) , 10.1098/RSTA.1984.0069
S. Skyum, L. G. Valiant, A complexity theory based on Boolean algebra Journal of the ACM. ,vol. 32, pp. 484- 502 ,(1985) , 10.1145/3149.3158
Peter E. Hart, Richard O. Duda, Pattern classification and scene analysis A Wiley-Interscience Publication. ,(1973)