AGRICULTURAL PRODUCTIVITY FORECAST IN RURAL SETTLEMENTS IN MATO GROSSO DO SUL USING ARTIFICIAL INTELLIGENCE

Abstract

t is very important to be able to predict the production of agriculture and livestock in the settlements of a region, so that public policies are aimed at improving the life of its settlers. The objective of this study was to predict the productivity of agriculture in rural settlements in Mato Grosso do Sul (MS). In agriculture, corn, beans and cassava were chosen, and in livestock, bovine, swine and poultry were chosen. The sample consisted of twenty-six rural settlements in MS, collecting socioeconomic and demographic data from the settlers and their production in agriculture and livestock. The variables studied in relation to agriculture were the production of corn, beans and cassava and, in relation to livestock, bovine, swine and poultry. With the averages of these productivities and, in possession of the levels of productivity of each variable in MS provided by specialists in the subjects, artificial intelligence (fuzzy logic) was applied to facilitate the decisions about the agricultural and livestock production of these rural settlements, about those data that were between the completely true and the completely false. The results obtained showed average productivity both in the agricultural and livestock sectors of rural settlements in MS.