论文信息 - Study on optimization of support vectors

Study on optimization of support vectors

A novel method on the reduction of support vectors in the support vector machine (SVM) is presented. The reduced support vectors in the SVM are applied to the agent initial position optimization of the land combat simulation model. Numerical simulations show that the running efficiency could be increased more then 370 times employing the proposed method compared with that using the conventional SVM method under the original precision. Some experiences and trends in the study on the optimization problem are also summarized and presented.

Yanchun Liang | H. P. Lee | L. M. Wan | C. G. Wu

[1] John C. Platt,et al. Fast training of support vector machines using sequential minimal optimization, advances in kernel methods , 1999 .

[2] H. P. Lee,et al. COMBINING SUPPORT VECTOR MACHINE (SVR) WITH GENETIC ALGORITHM (GA) TO OPTIMIZE THE INITIAL POSITIONS OF AGENTS IN THE LAND COMBAT SIMULATION , 2002 .

[3] Vladimir N. Vapnik,et al. The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[4] Chunguo Wu,et al. A computational-intelligence-based optimization of agents' initial positions in land combat simulation , 2003 .

[5] Kimberly A. Roddy,et al. Modeling Human and Organizational Behavior Using a Relation-Centric Multi-Agent System Design Paradigm (with) RELATE: A Relation-Centric MAS Simulation Development Package (CD-ROM) , 2000 .