Skip to Main Content
The paper is mainly about the application of Particle Swarm Optimization (PSO) algorithm to specify parameters of Support Vector Machine (SVM). In this paper, sparseness of SVM's solution was introduced to improve fitness function of PSO algorithm. Summation of empirical risk and count of support vectors divided by training set's size was employed as fitness function. Simulation results proved that the improved PSO-SVM algorithm avoided “over-fitting problem” in parameter optimization process, and prediction precision of SVM was guaranteed.
Intelligent Human-Machine Systems and Cybernetics (IHMSC), 2010 2nd International Conference on (Volume:2 )
Date of Conference: 26-28 Aug. 2010