A Learning Process Using SVMs for Multi-agents Decision Classification.

In order to resolve decision classification problem in multipleagents system, this paper first introduces the architecture ofmultiple agents system. It then proposes a support vector machinesbased assessment approach, which has the ability to learn the rulesform previous assessment results from domain experts. Finally, theexperiment are conducted on the artificially dataset to illustratehow the proposed works, and the results show the proposed method haseffective learning ability for decision classification problems.