中文摘要：

随着互联网＋时代的到来，有许多公司依托互联网建立了打车软件服务平台，以实现乘客与出租车司机之间的信息互通，但是由于乘客与出租车司机信息不匹配，导致“打车难”的现象时有发生．本文首先建立了基于主成分分析和聚类分析的出租车资源“匹配程度”模型．其次在建立回归分析的基础上，构建二元回归模型，通过实证分析，得出的结果验证了打车补贴政策一定程度上确实缓解了打车难的模型．最后，由于不同时空下的供求匹配程度不同，我们主要建立了针对出租车司机的最优补贴模型，对出租车资源配置优化给出了一些相关建议．

英文摘要：

With the advent of Intemet ＋ times, many companies has established the taxi - taking software service platform based on the interact to realize the information connection between passengers and taxi drivers. However, due to the mismat- ches of the information betweenthem, the phenomena of being difficult to take a taxi occur occasionally. In this essay, firstly , we construct a model of the ＂matching degree＂ of taxi resources based on the principal components analysis and cluster analy- sis. Secondly, a binary regression model is madebased on the regression analysis. The results produced from the empirical a- nalysis prove that the subsidy policy in taxi taking has released the difficulties to take a taxi to some extent. Finally, due to the different matching degree of the supply and demand at different time and places, we build an optimal subsidy model for the taxi drivers, and propose some suggestions to optimize the taxi resources allocation.