There are difficult to quantify institutional factors in the development model of agricultural commercialization, which forms the measurement deviation of trade efficiency, resulting in large estimation relative error. To solve this problem, this paper studies the development of agricultural commercialization from the perspective of trade equality based on data mining algorithm. Preprocess agricultural multi-source heterogeneous data, remove abnormal data, establish agricultural commercialization database, input agricultural data into the database according to attributes and types, and realize the integration of multi-source data. Set the appropriate minimum support and confidence, mine the export volume of agricultural products from the perspective of high-level and low-level, and determine the structure of trade products. On the basis of using data mining algorithm to grasp the structure of agricultural products, an estimation model of agricultural commercialization development potential is established. The experimental results show that the average relative error of this method is 0.523, which is 0.453 and 0.324 lower than that based on BP neural network and RBF neural network. Therefore, the estimation results of this method are more accurate, can truly reflect the trade development capacity of agricultural products, and has a good application prospect.
Read full abstract