Application of Machine Learning in Comprehensive Evaluation of Agricultural High-tech

Under the background of the national policy of promoting agriculture through science and technology, agricultural high-tech are emerging continuously. How to evaluate these technologies objectively and efficiently to help enterprises make decisions is of great significance to promote the industrialization of technologies. To solve problems in traditional evaluation process,such as the lag of evaluation method, the subjective influence of experts, time consuming, labor consuming and low efficiency, an intelligent comprehensive evaluation method based on machine learning was proposed. To establish a hierarchical evaluation index system using Analytic Hierarchy Process (AHP), the index system is firstly quantified, and then 25 agricultural high-tech in different fields are selected and scored by experts; To build BP Neural Network model, the data obtained from experts' experience are selected as samples to train the model, achieve the intelligent comprehensive evaluation and calculation of agricultural high-tech. The experiment proves that compared with the Support Vector Machine(SVM) model, the absolute error between the output value and the real value obtained by BP Neural Network(BP-NN) model is within the range of (-0.06~0.08), the average relative error is 0.0265%. It can be seen that BP Neural Network applied to the comprehensive evaluation of agricultural high-tech is with high accuracy and simple usage, and can meet the requirements of agricultural high-tech comprehensive evaluation.