Research on LSSVM based on analytic selection and simulate anneal arithmetic

DOI

10.2991/icmmcce-15.2015.208How to use a DOI?

Keywords

Software Reliability Model; Least Square Support Vector Machines (LSSVM); Analytic Selection (AS); Simulate Anneal Arithmetic (SA)

Abstract

In view of the existing situation of artificial parameter selection in the fitting process which using the LSSVM software reliability model, the combination method of analytic selection (AS) and simulate anneal arithmetic (SA) is proposed for parameter selection, then obtain the improved software reliability model based on ASSA-LSSVM. The ASSA-LSSVM model considers the optimization of the penalty parameters C and kernel parameter in the model, make the model more realistic, and reduce the time cost of artificial selection-parameter. Make use of the published three groups of software failure data to verify this proposed model, the result shows that, compared with the former LSSVM model, the ASSA -LSSVM model can achieve better fitting results.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Dan Yang
AU  - Lie Zang
AU  - Lu Gan
PY  - 2015/12
DA  - 2015/12
TI  - Research on LSSVM based on analytic selection and simulate anneal arithmetic
BT  - Proceedings of the 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering 2015
PB  - Atlantis Press
SP  - 964
EP  - 970
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmmcce-15.2015.208
DO  - 10.2991/icmmcce-15.2015.208
ID  - Yang2015/12
ER  -