An Integrative Approach Towards Recommending Farming Solutions for Sustainable Agriculture

Abstract

Sustainable Agriculture is rapidly emerging as an important discipline to meet societal needs for food and other resources by adopting paradigms of conserving natural resources while maximizing productivity benefits. This paper proposes an integrative methodological approach for critically analyzing Precision Farming (PF) paradigms and Zero Budget Natural Farming (ZBNF), providing sustainable farming solutions and achieving productivity and profitability. This paper analyses the productivity of crops in PF using various machine learning (ML) algorithms based on different soil and climatic factors to identify sustainable agricultural practices for maximizing crop production and generating recommendations for the farmers. When implemented on the collected dataset from various Indian states, the Random Forest (RF) model produced the best results with an AUC-ROC of 95.7%. The Juxtaposition of ZBNF and non-ZBNF is evinced. ZBNF is statistically (p<0.05) observed to be a cost-efficient and more profitable alternative. The impact of ZBNF on soil microbial diversity and micro-nutrients is also discussed.