Automation in Agriculture: A Systematic Survey of Research Activities in Agriculture Decision Support Systems Using Machine Learning

Abstract

AbstractIn this age of automation, Machine learning (ML) plays the main role in agriculture sector to suggest suitable advice, crop advice, which includes decisions of growing crops, and advice related to growing season for precision farming. This systematic literature review performs a review of 103 documents of different ML approaches to analyze the performance of algorithms and used features in the work of prediction of crop yield and decision support systems to solve agriculture problems. These 103 documents are retrieved from different electronic databases, for analysis. The paperwork presents methods, accuracy measures, and used agriculture parameters, to understand the existing work done by authors. According to analysis, most of the authors used N, P, and K values and type of soil, and most of the authors used classification techniques such as Support Vector Machine, Decision Trees, Regression techniques, Random Forest, and Naive Bayes algorithm; the most applied clustering algorithm in the existing work is K-means. As per the additional survey, the Convolution Neural Network (CNN) algorithm is used by most of the authors for image processing in their work. Also, survey shows that very few authors used associative classifiers and association rule mining techniques to solve the agriculture problems.KeywordsAssociative classifierAgricultureClassificationAssociation rule miningClusteringNeural networkDecision support systemMachine learningConvolution neural network