A DATA-DRIVEN APPROACH TO CROP YIELD PREDICTION USING ADVANCED MACHINE LEARNING TECHNIQUES

Abstract

Half or more of India's population relies on agriculture for their livelihood, making it an essential sector of the Indian economy. The future of agriculture is in jeopardy due to the growing threat posed by climate change and other environmental factors. Improving decision-making about agricultural cultivation and growing practices, machine learning (ML) offers a tool for crop yield prediction (CYP). Several approaches have been devised to analyze AI-based crop yield prediction algorithms; this study centers on a systematic review that extracts and synthesizes CYP features. Less relative inaccuracy and less capacity to predict crop yield are the primary drawbacks of neural networks. Supervised learning algorithms had a hard time selecting, sorting, or rating fruits due to the nonlinear connection between input and output variables. Many agricultural development research proposals sought to build an accurate and efficient model for crop classification, which would allow for the prediction of crop yields in response to weather and disease conditions, the categorization of crops according to their developmental stage, and so on. An extensive evaluation of the accuracy of various machine learning models used to estimate agricultural productivity is presented in this article.