Construction of class regions by a randomized algorithm: A randomized subclass method

被引：25

作者：

Kudo, M

Yanagi, S

Shimbo, M

机构：

[1] Dept. of Information Engineering, Faculty of Engineering, Hokkaido University

来源：

PATTERN RECOGNITION | 1996年 / 29卷 / 04期

关键词：

subclass method; randomized algorithm; discrimination; hyper-rectangles; PAC learning; computational learning theory;

D O I：

10.1016/0031-3203(95)00107-7

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 [模式识别与智能系统]; 0812 [计算机科学与技术]; 0835 [软件工程]; 1405 [智能科学与技术];

摘要：

A randomized algorithm is proposed for solving the problem of finding hyper-rectangles, sufficiently approximating the true region in each class. This method yields a suboptimal solution, but is more efficient than previous methods. The performance is analysed based on a criterion of PAC (Probably Approximately Correct) learning. Experimental results show that the proposed method can solve large problems which were not able to be solved previously.

引用

页码：581 / 588

页数：8

共 12 条

[1]

Glucksman H. A., 1967, 1ST ANN IEEE COMP C, P138

[2]

ICHINO M, 1989, IEICE J, P49

[3]

ICINO M, 1979, IEEE T SYST MAN CYB, V9, P345

[4]

OPTIMAL SUBCLASSES WITH DICHOTOMOUS-VARIABLES FOR FEATURE-SELECTION AND DISCRIMINATION [J].