A tool for classification of cacao production in Colombia based on multiple classifier systems

Abstract

Cacao is one of the central crops that support the agrarian production in Colombia. In some areas, it is the main source of income for about 25.000 families. Frequently, its production is affected by several factors such as climate, soil, water, wind, among others. This paper presents a Machine Learning approach for classifying cacao production in the region of Santander, Colombia. The proposed system aims to link climate and cacao production data to develop the classification task. In this sense, several techniques were experimentally evaluated in order to determine the algorithm that generates the best model to classify new climate instances on the cacao production dataset. Experimental results showed a better precision for Random Forest in comparison with other evaluated techniques.