The Research on the Algorithm of Approximately Linearly Dividable Support Vector Classification Machine Based on Fuzzy Theory

This paper presents a comprehensive review of computational Intelligence (CI) technology applied in structural damage identification, clarifies the basic principles of computational intelligence techniques, as well as the applicable difficulties that exist in the field of structural damage identification (SDI) from 6 aspects: fuzzy theory, evidence theory, rough set theory, artificial neural networks, support vector machines and evolutionary computation, and then discussed the applicable prospects of computational Intelligence in SDI. It points out that the reasonable cross-fusion of a variety of CI method to specific research object is a necessary means for SDI research. For economy and practicality considerations, CI is suitable for highly integrated complex structural damage identification.

Download Full-text

Analysis of Speaker Verification System Using Support Vector Machine

JOURNAL OF ADVANCES IN CHEMISTRY ◽

10.24297/jac.v13i10.5839 ◽

2017 ◽

Vol 13 (10) ◽

pp. 6531-6542

Author(s):

P Shanmugapriya ◽

Y. Venkataramani

Keyword(s):

Support Vector Machine ◽

Speaker Recognition ◽

Speaker Verification ◽

Fuzzy Theory ◽

Support Vector ◽

Fuzzy Support Vector Machine ◽

System A ◽

Verification System ◽

Svm Model

The integration of GMM- super vector and Support Vector Machine (SVM) has become one of most popular strategy in text-independent speaker verification system.Â This paper describes the application of Fuzzy Support Vector Machine (FSVM) for classification of speakers using GMM-super vectors. Super vectors are formed by stacking the mean vectors of adapted GMMs from UBM using maximum a posteriori (MAP). GMM super vectors characterize speakerâ€™s acoustic characteristics which are used for developing a speaker dependent fuzzy SVM model. Introducing fuzzy theory in support vector machine yields better classification accuracy and requires less number of support vectors. Experiments were conducted on 2001 NIST speaker recognition evaluation corpus. Performance of GMM-FSVM based speaker verification system is compared with the conventional GMM-UBM and GMM-SVM based systems.Â Experimental results indicate that the fuzzy SVM based speaker verification system with GMM super vector achieves better performance to GMM-UBM system. Â

Download Full-text

110 Detailed Diagnosis for Rotating Machinery Using Support Vector Machine and Fuzzy Theory

The Proceedings of the Symposium on Evaluation and Diagnosis ◽

10.1299/jsmesed.2011.10.39 ◽

2011 ◽

Vol 2011.10 (0) ◽

pp. 39-43

Author(s):

Hongtao Xue ◽

Yutaka Jin ◽

Peng Chen

Keyword(s):

Support Vector Machine ◽

Fuzzy Theory ◽

Rotating Machinery ◽

Support Vector ◽

Detailed Diagnosis

Download Full-text

Fault Diagnosis of Conveyance Machine Based on Fuzzy Support Vector

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.135-136.547 ◽

2011 ◽

Vol 135-136 ◽

pp. 547-552

Author(s):

Yuan Bin Hou ◽

Ning Li ◽

Fan Guo ◽

Jing Chen

Keyword(s):

Support Vector Machine ◽

Fault Diagnosis ◽

Kernel Function ◽

Coal Mine ◽

Fuzzy Theory ◽

Kernel Functions ◽

Mine Safety ◽

Support Vector ◽

Coal Mine Safety ◽

Rubber Belt

Aiming random and nonlinearity for conveyance machine of rubber belt in mine, a method of fault diagnosis is presented which fusion of fuzzy theory and support vector machine (FSVM). According to the coal mine safety rules of the regulation, the conveyance machine servicing are deduced eleven faults after analyzing practice statistic data, here, we consider some are fuzzy that the statistic data are divided to the normal kind or fault kind, but some are definite that the statistic data possibility are belong to same kind fault, accordingly, the fuzzy support vectors is established. Farther, two kernel functions of FSVM is made for seeking the problem of random and nonlinearity, which are RBF and TANH. According to the random statistic data and the study sample, analyzing the effect of expense and kernel function in selecting different parameters, the unitary constant is ascertained, next, the FSVM kernel function of fault diagnosis multi-class rules are established, then, this method availability is proved using test data and simulation.

Download Full-text