EFFECTIVE PREDICTION USING ENSEMBLE MACHINE LEARNING TECHNIQUES FOR THE TOMATO CROP RECOMMENDATION SYSTEM

Abstract

Agriculture is the fundamental industry responsible for food production and supplying the
necessary raw materials for other industrial activities. Agricultural production growth isn't keeping
pace with population expansion, which might lead to a global food shortage. Thus, developing
nations with limited resources and territories must produce more food. Choosing a regionally
appropriate agricultural product boosts productivity. Past data on environmental conditions,
cultivation areas, and tomato crop output quantities are needed to anticipate agricultural production
in a region. The data used to make these forecasts is private. Since India is a growing country with
an agrarian economy, this dissertation focuses on it. We start by acquiring and preparing relevant
Indian Agriculture and Welfare Department data. Subsequently, we introduce a sophisticated
ensemble machine learning approach known as Multi-layer Perception Random Forest Regression
(MLPRFR) to precisely forecast the yield of the primary crops, namely wheat, tomato, banana,
and rice.