Title :
ε-Insensitive Modification of Subspace Information Criterion
Author_Institution :
Fac. of Math. & Inf. Sci., Huanggang Normal Univ., Huanggang, China
Abstract :
Evaluating the generalization performance of learning machines without using additional test samples is one of the most important issues in the machine learning community. The subspace information criterion (SIC) is one of the methods for this purpose, which is shown to be an unbiased estimator of the generalization error with finite samples. In this paper, we give ε-insensitive modification of the subspace information criterion (mSIC), it can improve the precision of SIC.
Keywords :
generalisation (artificial intelligence); learning (artificial intelligence); ε-insensitive modification; generalization error; learning machine; performance evaluation; subspace information criterion; Degradation; Function approximation; Genetics; Information science; Kernel; Machine learning; Mathematics; Parameter estimation; Silicon carbide; Testing; generalization error; insensitive modification; precision; subspace information criterion;
Conference_Titel :
Genetic and Evolutionary Computing, 2009. WGEC '09. 3rd International Conference on
Conference_Location :
Guilin
Print_ISBN :
978-0-7695-3899-0
DOI :
10.1109/WGEC.2009.60
