基于PCA的SVM算法在网络欺诈行为中的研究与应用

The Research and Application of SVM Algorithm Based on PCA in Network Fraud

随着信息时代的蓬勃发展,网络订单无序的爆炸式膨胀,交易数据可能出现的信息失真使得快速交易的风险呈指数增长,导致越来越多的网络欺诈行为被推向了风口浪尖。本文通过分析网络机票订单的历史数据,采用主成分分析(Principal Component Analysis,PCA)的方法对其特征空间进行降维,运用支持向量机(Support Vector Machine,SVM)算法进行模型训练,从数据和模型层面得到分类预测模型,从而实现对网络订单欺诈行为的风险预测和监控。

With the prosperous of Internet Times, the inflation it's no arrangements to online ordering and the information distortion that the transaction data may appear makes the risk of the rapid growing exponentially, which promote more and more online fraud to the top. The assignment, according to analysis the historic data of online flight and the feature space is reduced by the method of principal component analysis, and use model training to support vector machine algorithm.And we receive the sorts based on data and model, thus to realize the prediction and monitoring of the online fraud.