Research On Logistics Evaluation Based On Entropy Weight Method-bp Neural Network

With the continuous growth of China's economy,the modern logistics industry has broad market prospects.It is an urgent need that people evaluate the development of regional logistics objectively and comprehensively to lead the corporation decision-making and government management.However,the regional logistics evaluation in China is still in its infancy,the index system is not comprehensively constructed,and the evaluation method is subjective.As a result,in order to be able to quantify and evaluate the logistics accurately,this article takes Jiangsu Province as an example to construct a set of regional logistics index models,which are conducive to the government's macro management and promote regional economic growth.Taking the regional logistics evaluation as the research object,this paper first analyzes the advantages and disadvantages of logistics index construction and method evaluation at home and abroad.Secondly,taking the relevant data of Jiangsu Province from 2002 to 2018 as an example,following the principles of accessibility,representativeness and independence of indicator selection,the initial indicator system is constructed.Then,it quantitatively screen regional logistics indicators through the R-type clustering method of class average method and coefficient of variation.The selected regional logistics index system is mainly composed of five secondary indicators and 25 tertiary indicators including basic logistics index,development potential index,logistics benefit index,environmental impact index and economic development index.Then,an evaluation model based on the entropy weight method-BP neural network is proposed,namely,the initial index weight is obtained by using the entropy weight method,and the initial regional logistics comprehensive index is obtained by weighting.The standardized initial index value and initial regional logistics comprehensive index value are used as the input of the BP neural network,and the final index weight is the output.The initial evaluation results of the entropy weight method are divided into 16 training samples and 1 test sample.Through matlab 2016 a simulation,the weighted Jiangsu comprehensive logisticsindex and each sub-logistics index are obtained.Finally,the trend and cause of the total regional logistics index and each sub-logistics index of Jiangsu Province are analyzed by line charts and pie charts.The paper points out that the government should pay attention to the management of railway,pipeline transportation and the training of logistics personnel,give certain support to logistics enterprises,pay attention to ecological protection,and control pollution emissions,thereby improving the logistics economy of Jiangsu Province...