TY - GEN
T1 - Learning performance of DAGSVM algorithm based on Markov sampling
AU - Xu, Jie
AU - Lu, Yang
AU - Zou, Bin
N1 - This work is supported in part by NSFC project (61370002, 61403132). The corresponding author is Bin Zou.
Publisher Copyright:
© 2015 IEEE.
PY - 2015/7/12
Y1 - 2015/7/12
N2 - Support vector machine (SVM) is originally designed for 2-class classification problem under the assumption of independent and identically distributed (i.i.d.) sampling. Most classification problems in practice involve multiple categories, hence the SVM has been extended to handle multi-class classification by solving a series of binary classification problems such as the Directed Acyclic Graph SVM (DAGSVM) method. In this paper, we propose the new DAGSVM based on the Markov sampling to replace the classical framework of i.i.d. samples. Numerical studies on the learning performance of the DAGSVM based on Markov sampling for real-world dátasete are presented. Experimental results indicate that the DAGSVM based on Markov sampling yields better learning performance compared to the DAGSVM algorithm based on independent random sampling.
AB - Support vector machine (SVM) is originally designed for 2-class classification problem under the assumption of independent and identically distributed (i.i.d.) sampling. Most classification problems in practice involve multiple categories, hence the SVM has been extended to handle multi-class classification by solving a series of binary classification problems such as the Directed Acyclic Graph SVM (DAGSVM) method. In this paper, we propose the new DAGSVM based on the Markov sampling to replace the classical framework of i.i.d. samples. Numerical studies on the learning performance of the DAGSVM based on Markov sampling for real-world dátasete are presented. Experimental results indicate that the DAGSVM based on Markov sampling yields better learning performance compared to the DAGSVM algorithm based on independent random sampling.
KW - Directed a-cyclic graph SVM (DAGSVM)
KW - Learning performance
KW - Markov sampling
KW - Multi-class classification
UR - http://www.scopus.com/inward/record.url?scp=85020696342&partnerID=8YFLogxK
U2 - 10.1109/ICMLC.2015.7340674
DO - 10.1109/ICMLC.2015.7340674
M3 - Conference proceeding
AN - SCOPUS:85020696342
T3 - Proceedings - International Conference on Machine Learning and Cybernetics
SP - 910
EP - 915
BT - Proceedings of 2015 International Conference on Machine Learning and Cybernetics, ICMLC 2015
PB - IEEE
T2 - 14th International Conference on Machine Learning and Cybernetics, ICMLC 2015
Y2 - 12 July 2015 through 15 July 2015
ER -
