WSEAS Transactions on Mathematics
Print ISSN: 1109-2769, E-ISSN: 2224-2880
Volume 13, 2014
NLS-TSTM: A Novel and Fast Nonlinear Image Classification Method
Authors: , ,
Abstract: This paper is devoted to study nonlinear image classification methods based on support tensor machine (STM). Firstly, a new linear method named linear least squares twin support tensor machine (LLS-TSTM) is proposed, which is an improvement of linear STM. The utility of twin skill and least squares technology aims to speed up the computation time (sum of training time and testing time). Secondly, in order to study nonlinear version of LS-TSTM, a new matrix kernel function is introduced and then based on which, a nonlinear LS-TSTM (NLS-TSTM) classification method is suggested with detailed theoretical derivation. Finally, in order to examine the effectiveness of LLS- and NLS-TSTM, we perform a series of comparative experiments with linear STM and linear TSTM on ORL and Yale face databases. Experiment results show that the proposed methods are effective and efficient.
Search Articles
Keywords: Least squares twin support tensor machine, image classification, matrix kernel function, iterative algorithm, classification accuracy
Pages: 626-635
WSEAS Transactions on Mathematics, ISSN / E-ISSN: 1109-2769 / 2224-2880, Volume 13, 2014, Art. #61