Learning Pattern Classification—A Survey英文文献资料.pdf

2178 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 44, NO. 6, OCTOBER 1998 Learning Pattern Classification—A Survey Sanjeev R. Kulkarni, Senior Member, IEEE, Gabor Lugosi, and Santosh S. Venkatesh, Member, IEEE ´ (Invited Paper) Abstract— Classical and recent results in statistical pattern tern classification, stochastic control, reinforcement learning, recognition and learning theory are reviewed in a two-class clustering, etc. pattern classification setting. This basic model best illustrates Work on these areas spans a number of fields and many intuition and analysis techniques while still containing the essen- tial features and serving as a prototype for many applications. years. In the last several decades, there was work in the Topics discussed include nearest neighbor, kernel, and histogram 1940’s and 1950’s in areas such as statistics, information methods, Vapnik–Chervonenkis theory, and neural networks. The theory, cybernetics, and early work on neural networks that presentation and the large (thogh nonexhaustive) list of references led to tremendous progress and, in fact, established several is geared to provide a useful overview of this field for both new fields of activity. Continued work in these areas, slightly specialists and nonspecialists. later work on systems and control theory, pattern recognition, Index Terms—Classification, learning, statistical pattern recog- and optimization, and more recently th