Application Research Of Support Vector Machine In Port Logistics Demand Analysis

Port logistics plays a more and more important role in the economic development of our country.From the point of view of port city,port logistics is more and more important to the development of port city and its economic hinterland.From the development of the port to domestic and foreign experience,the development of port logistics can effectively promote the coordinated development between city and port economic hinterland,and economic hinterland fast in turn for port logistics to provide a stable supply,the two are complementary to each other.At present,China's port logistics in a period of rapid development,the major ports are set to system planning and construction of port logistics,one of the key part is the analysis of port logistics demand,the port logistics needs analysis is the premise and foundation of making development strategy of port logistics,port logistics is an important basis for the functional orientation,space the layout and development direction of comprehensive planning.In this paper,through the analysis of the internal drive between the development of port logistics and the hinterland economy,according to the influencing factors of port logistics demand,and in accordance with the principle of effectiveness,the indexes of the relative independence principle,operational principle,strong correlation principle,established port logistics demand analysis quantitative index set,and then through the grey correlation analysis method,select the specific indicators for port logistics demand analysis of Shenzhen city;then,based on the domestic and overseas were analyzed and compared,based on the combination of principal component analysis and support vector machine the advantages of two algorithms based on PCA-SVM model is constructed in this paper as a prediction model of port logistics demand scale,and in Shenzhen as an example for the empirical analysis;then from the port logistics demand structure and main body,flow pattern,demand regulation The four aspects of Shenzhen port logistics demand analysis and forecast,the results show that in 2016 Shenzhen port throughput of 213 million 283 thousand and 200 tons,compared with last year is expected to fall by about 1.74%.The main influencing factors were found out by sensitivity analysis,and the future development of Shenzhen port was put forward.This paper believes that the development of port logistics demand and the change of the economic hinterland is closely related in research and analysis,not only rely on the historical data of port logistics demand,to combine between the economic hinterland of the port,establish scientific and reasonable quantitative index set.In the selection of prediction methods,we should choose the prediction method according to the characteristics of port logistics demand,the paper selects the forecasting support vector machine has a good applicability analysis as the analysis model in small samples,nonlinear,provides better research ideas and perspectives for the analysis and research of port logistics demand.