An intelligent model selection scheme based on particle swarm optimization

Author

Huang, Jingtao ; Chi, Xiaomei ; Ma, Jianwei

Author_Institution

Electron. & Inf. Eng. Coll., Henan Univ. of Sci. & Technol., Luoyang, China

Volume

1

fYear

2009

fDate

20-22 Nov. 2009

Firstpage

882

Lastpage

886

Abstract

To improve the learning efficiency of support vector machine, an intelligent model selection scheme based on particle swarm optimization (PSO) was presented to optimize the hyper-parameters. By taking the model selection problem as a multi-object optimization problem, one can obtain a solution set known as Pareto front; each one model in this set is non-dominated. PSO was used to solve the above multi-objective optimization problem and then the model set was obtained. The scheme was tested on several datasets, the results show that Pareto front can be obtained in one trial and the effect of every single parameter can be displayed more directly.

Keywords

learning (artificial intelligence); particle swarm optimisation; support vector machines; Pareto front; hyperparameters; intelligent model selection; learning efficiency; multiobject optimization problem; multiobjective optimization; particle swarm optimization; support vector machine; Computer errors; Educational institutions; Learning systems; Machine intelligence; Machine learning; Pareto optimization; Particle swarm optimization; Statistical learning; Support vector machine classification; Support vector machines; Pareto front; intelligent model selection; multi-object optimization; particle swarm optimization(PSO); support vector machine;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Computing and Intelligent Systems, 2009. ICIS 2009. IEEE International Conference on

Conference_Location

Shanghai

Print_ISBN

978-1-4244-4754-1

Electronic_ISBN

978-1-4244-4738-1

Type

conf

DOI

10.1109/ICICISYS.2009.5358047

Filename

5358047