Experimental evaluation of expert fusion strategies

被引:63
作者
Alkoot, FM [1 ]
Kittler, J [1 ]
机构
[1] Univ Surrey, Sch Elect Engn Informat Technol & Math, Ctr Vis Speech & Signal Proc, Guildford GU2 5XH, Surrey, England
关键词
classifier combination; posterior probability; estimation accuracy; error distribution;
D O I
10.1016/S0167-8655(99)00107-5
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We investigate the classifier combination models presented in (Kittler et al., 1998; Kittler, 1998) and validate them experimentally. We emulate the behaviour of individual experts by subjecting their nominal soft outputs to perturbation errors. A relation between the effectiveness of various combination strategies and estimation errors is established. (C) 1999 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:1361 / 1369
页数:9
相关论文
共 19 条