Vehicle Type Recognition Based On Data Mining Research

Data Mining is a very important research rising in IT field in recent years. In transportation area, a large number of source data’s value of numerous vehicles lies in the lots of unknown information, which can provide people with more help. With Data Acquisition Technology has been developed rapidly, in the meantime, data volume has become larger and attributive character has increased more, therefore, the essential problem of data mining is that how to acquire useful information from these data.In recent years, with the booming of social science and technology and world economy, transport industry act as an important field which relate to people’s daily life has become one of the focus. Since rapid growth of the quantities of vehicles in nowadays, the problem which require immediately solution is that how to have a better understanding of various vehicle’s situations to conduct real-time monitoring by existing information; how to identify and classify vehicles in specified area and time, and use it as a basis of traffic control, fee charging, dispatching and statistics, thus realize speedy traffic control. Aiming at above issues, this article analyzes the significance of vehicle’s type classification and the insufficient of traditional manual classification, and proposes that Applied Data Mining method has practical value in identification of vehicle’s classification, transport industry and management industry.This paper aims at the available data, modeling and forecasting with classification method. First of all, this paper introduces several methods of Data Mining and the characteristics of each model by which they constructed, then applys them to vehicle identification to judge their identification ability and proceed comparative analysis. Deriving from the principal component analysis and discriminant analysis, these two types "bus" and "van" in four categories has high purity of identification in training sets and testing sets, therefore the four categories converted into two categories focusing on dealing with opel category and saab category. After processing the source data, finally compare the results of each model.