Application Of Data Mining In Urban Road Traffic

Along with the advancement of modernization, the expansion of city scale, the management of the modern city is facing more and more difficult, and the urban traffic is a city economy and an important part of urban culture, from the earliest formulation,intelligent transportation to solve the traffic problems, the intelligent transportation to alleviate the traffic problem, now the wisdom city traffic solution, the study of the urban road traffic has become one of the focuses of current city solution, and also a focus of research and application. Among them, the large-scale urban transportation information integration, analysis, management is a key technology and mining, and data mining technology is one of the important data to solve the urban road traffic technology, it can easily deal with huge amounts of data and based on the data for the traffic rule in the data, to provide technical support for the intelligent traffic solutions, is advantageous to the scientific and efficient management of urban traffic, ease traffic congestion, illegal traffic behavior, optimize the transportation network, promote the healthy development of urban traffic management to provide effective solutions for the city.The purpose of this paper is to explore the practical application of data mining on the road traffic, first analyzed the research status quo of urban road traffic and the existing problems, reveals the data mining technology in the field of urban road traffic status. Secondly, this paper introduces data mining technology application the main direction of urban road traffic and the corresponding data mining algorithm, and successively to explore the analysis of the road traffic flow forecast, road traffic factors in mining, road traffic flow distribution pattern mining, integrated urban road intelligent transportation system and so on four main direction.Then, in view of the deck in the study of urban road traffic vehicle mining problem,collecting relevant data, USES the relatively advanced and efficient data processing tool Hadoop and data warehouse tool Hive, and the various tools do a simple introduction.Aimed at the quality of the data made in-depth analysis, based on the analysis of the visual display means make more intuitive and comprehensive analysis of the problem,and according to the business logic to design a set of relatively complete data cleansing rules.Finally, in-depth examples, this paper has studied three kinds of mining algorithm suspect vehicle deck, respectively is based on the probability to bayonet deck vehiclesmining algorithm, based on the weighted average travel time of mining algorithm and integrated deck vehicle and improve the first two algorithm is proposed based on the geographic space and time of deck vehicle comprehensive mining algorithm, the feasibility of the algorithm is analyzed in theory, and it has been proved that after the two algorithms is overcome and improve the former comprehensive mining algorithm indeed has the better effect of mining and mining accuracy, completed a comprehensive mining algorithm in the theory and examples on the scientific nature and effectiveness of proof.