A predictive model construction for mulberry crop productivity.

Abstract

India accounts for more than fifty percent of sericulture production in the world. The modern Sericulture methods that have evolved demand, accurate classification of soil suitable for Mulberry crop productivity. But the most prevalent method adopted currently in soil testing is manual, which often fails to give the correct prescription to make soil suitable for Mulberry crop. A scientific approach of soil testing could aid farmers in dynamic decision-making, which would significantly increase Mulberry crop productivity. Such analysis is possible with the help of data analysis, thanks to the advent of modern computer technology. Due to significant advances in the area of Information Technology and agriculture, there is scope of interdisciplinary work, application thereof to solve agricultural problems. Hence effort was made to explore and develop an automated system for the analysis of range of soil characteristic suitable for Mulberry crop production, which in turn contribute to increase in Cocoon productivity. The experiment was carried out by collecting soil samples from different irrigated regions of Karnataka, India, to deduce the range of soil parameters supporting the healthy growth of Mulberry crop. Further, different classification technique was applied on parameters of soil suitable for Mulberry crop using Hunt’s algorithm, and J48 Decision tree was more applicable in decision making. The statistical information obtained from data mining technique were validated through mathematical model for developing a forewarning predictive system for crop productivity.