Research On Data Mining And Road Logistics Index Based On Logistics Vehicle

In the context of the vigorous development of the modern logistics industry,urban logistics has become a part of the urban competitiveness.At present,road logistics as the main mode of the urban logistics transportation.Through the study of the urban road logistics index,contribute to make a judgment to the development of the logistics industry and the general economy,to the benefit of enhancing further communication?comparison and research with the world developed logistics country.Along with the further applications of big data in the logistics,make a relevant analysis and processing to the logistics data,excavating useful information,in favor of the government or enterprises draw up a plan or make a decision,providing accurate data for reference.There is no specific previous research on road logistics,so the point of penetration based on the logistics enterprise data in this article,combining with the economic indicators of road logistics,building urban road logistics index.Logistics enterprise data mainly includes the waybill information data and the vehicle GPS data,however,information about waybill often involving trade secrets,and distributed to different inner-enterprise,not for publication.Based on this reason,this thesis is based on the logistics vehicles GPS data.Then,we can analyze and excavate data and lay a foundation for establishment of index.Firstly,take advantage of Hadoop platform,which is used to analyze off line data.We use it to analyze GPS data that logistics vehicles uploaded monthly.In order to mining out-degree?in-degree?the number of Active vehicle and the remaining time of the vehicle on a city,and on this basis,according to Google page rank PageRank algorithm,put forward the improved Urban Logistics Rank algorithm.Then calculating the urban logistics rankings.To a certain extent,ULR can reduce the error from web link phenomenon and Logistics Company's service network.Secondly,based on Freight statistics and some economic indicators from Shenzhen city bureau of statistics and Shenzhen traffic management committee.Urban road logistics index is obtained by principal component analysis(PCA)calculation.For purpose of verify the effectiveness of the index through comparing urban road logistics index with the city's main macroeconomic indicators using empirical analysis including Causality test.Thirdly,establish urban road logistics index prediction model,mainly based on RBF neural network model ? LS-SVM(least square support vector machine)regression prediction model,and PCA-LS-SVM combined forecasting model,which combined principal component analysis with the LS-SVM prediction model for urban road logistics index prediction.At last,providing a contrast and analysis to the prediction results.