Building a new model for feature optimization in agricultural sectors

Agriculture is an integral part of India. There are several factors that impact a healthy agricultural production. Enhancing the stability and growth of agricultural sector with the help of advanced technologies is a primary issue to be dealt with. Machine learning is an emerging area that can improve the agricultural productivity in India. In our research we have considered two vital agro datasets like mushroom and Soyabean. These datasets are subjected to PSO search algorithm that acts as an attribute selector. It reduces the datasets and removes the noisy features. To the resultant dataset different classifiers like Naive Bayes, Decision table are implemented and the final inference is deducted. The result was tested with some performance metrics like RMS and Classification accuracy. It is observed that Machine learning methods in combination with PSO search technique act as a positive factor in enhancing the overall performance.