Smart Agricultural Subsidy recommendation Approach for identifying appropriate formers using ML Techniques

Abstract

Subsidy on various types of items is more important in agriculture filed to optimize risk factors in production and also support to farmer to cultivate farming with extra benefits and with additional security. The traditional subsidy allocation technique heavily depends on physical verifications, checking large number of rules, verifying a greater number of constraints, these factors delayed the complete subsidy process that in turn facing the issue of identifying appropriate subsidy beneficiaries, and also shows inefficiency in utilizing government subsidy schemes. The farmers data samples are increased day by day with many features and this provide an opportunity to develop an intelligent technique to identify and allocate subsidies to needy farmers on time. In this research we propose an ML based SSR (Smart Subsidy Recommending) approach for allocation various kinds of subsidies, and the main objective is to develop efficient and transparent subsidy recommending techniques by analyzing complete farmer list. In our model, we have taken certain features like previous repayment histories, farmers economical situations, climatic conditions for crap, availability of farming tools, features of soil, characteristics of land, and farmer conditions to allocate various types of subsidies. The ML-SSR technique starts with preprocessing of farmers data, identification of feature vectors, remove irrelevant features with help of feature engineering approaches, allocation of data samples to various stages of ML, calculates various types of indices based on farmers data samples, and finally decide the list of eligible formers for different types of subsidies. Performance of ML-SSRS method is calculated and compared with RBEEA, EAA, ABSS, and TLR subsidy recommended approaches. The method we developed is useful to many government agencies to identify list of eligible farmers without any discrepancies and also prepare the list by maintaining transparency.