Current Trends and Applications of Machine Learning Techniques in Agriculture: An Insight

Abstract

Use of Machine Learning (ML) techniques for data-intensive agri technologies is a rapidly growing field of study. As a developing country, agriculture plays a vital role in the nation’s income. The main concerns in agriculture to enhance agricultural yield are to predict the right time to plant a crop, select best-suited soil, apply an appropriate amount of fertilizers, reduce the number of weeds, etc. to develop what is known as a smart irrigation system. If these factors are not properly taken care of, then crop quality, quantity, or productivity will get affected. Machine Learning approaches introduce precision and accuracy in the otherwise adopted techniques for enhancing agricultural yield. Numerous machine learning methods such as Support vector machines, Naïve Bayes, k-Nearest Neighbor, Support Vector Machines, and an Extreme Learning Machine classifier, Decision trees, Random forest, Neural network, Markov chain model, etc. The paper presents a comprehensive review of numerous Machine Learning approaches used in the subdomains of agriculture like soil management, fertilization, weed management, irrigation management, and early disease detection system for the crops. The in-depth study done elaborates the use of these approaches, their merits, demerits and also discusses the various challenges that are existing in the current state of research. The work also describes the various challenges faced by researchers in this domain. The paper concludes by analytically contrasting these machine learning techniques and describing which technique is more suited for a given type of soil.