Study On Location Selection Of Agricultural Products Logistics Distribution Center Based On Data Mining

The issue of "agriculture,rural areas and farmers" has always been the focus of attention in the course of China's economic development.With the continuous development of agriculture and the improvement of the level of science and technology,the output of agricultural products has greatly increased,which has accelerated the sales and circulation speed of agricultural products and greatly promoted the development of agricultural products logistics.Therefore,more and more attention has been paid to the logistics and distribution of agricultural products.At present,there are some problems in the logistics and distribution of agricultural products in our country,such as the high cost of logistics,low distribution efficiency,imperfect infrastructure,unreasonable location of distribution center and so on.In the whole logistics distribution,whether the location of distribution center is reasonable or not is very important to the whole logistics system,and how to solve this problem has become the focus of many scholars.The main content of this paper is to analyze and study the location-selection of agricultural products logistics distribution center,based on the domestic and foreign scholars' research,combined with data mining clustering algorithm and MapReduce model,to improve and optimize the location algorithm.First of all,from the perspective of the research background and significance of the paper and the current research situation of scholars at home and abroad,this paper puts forward the train of thought and framework of the paper.Secondly,it expounds the development and development model of agricultural products logistics and distribution in our country,and summarizes some problems in the current logistics of agricultural products in China.At the same time,the paper makes a comprehensive introduction and analysis of the location-selection of the agricultural products logistics distribution center,data mining technology,clustering algorithm and the basic concepts and related contents of MapReduce.Then,based on the classical algorithm clustering algorithm of data mining,the NK-Means clustering algorithm based on Dijkstra distance is proposed,and the NK-Means algorithm is optimized and improved with the MapReduce model to reduce the time complexity of the algorithm and improve the computing efficiency.Then,the algorithm is tested and analyzed through simulation cases.By taking the point coordinates and the demand as the main consideration,and comparing with the original clustering algorithm,the two algorithms are given different location-selection conditions under the same conditions,and the validity and superiority of the NK-Means algorithm are verified.Finally,the paper is summarized and prospect.