Ecological Agriculture Safety Supervision System Based on Random Forest Algorithm

The design and implementation of an ecological agriculture safety supervision system to complete the real-time collection, analysis and decision-making of agricultural information can not only greatly improve crop yield and food safety and quality, but also improve labor efficiency and economic benefits, and promote farmers increase in revenue. The purpose of this paper is to study the design of ecological agriculture safety supervision system based on random forest algorithm. Based on the difference of information flow process and system logic, we divide the ecological agriculture safety supervision system into four layers through a layered model: perception layer, transmission layer, control layer and application layer. Aiming at the shortcomings of random forest algorithm, an adaptive bagging random forest algorithm based on weight voting, ABW-RF algorithm, is proposed. The ABW-RF algorithm is applied to the intelligent assessment of soil heavy metal pollution risk. After preprocessing operations such as target value labeling of soil heavy metal data by the Nemerow comprehensive pollution index method, experiments were carried out on the soil data set in M area. The experimental results show that the recall rate and precision rate obtained by ABW-RF are the highest.