关键词: 物流工程, 应急物资, 后验概率支持向量机, 多类分类

Abstract: Once emergency happened,the requirements for materials are numerous and complicated, the requirements are urgent and variational as well in. To impove the efficiency of emergency logistics,the rescuer need to classify the emergency materials. A novel method for emergency materials classification based on posterior probability support vector machine is proposed. The classification model for emergency materials is presented via PPSVM. This article took the materials importance;urgency and scarcity as the classified criterion, put the velues after being fuzzed as the input of PPSVM, selected the classification which is corresponding with the Max Posterior Probability as the output , the results of the classification are obtained finally .Experimental results demonstrate that the proposed method is feasible, it can solve the problem of emergency materials classification effectively.

Key words: logistics engineering, emergency material;posterior probabilities support vector machine（PPSVM）;multi-class classification