A Fuzzy Classification System to Analyze the Yield of Cotton Production

Abstract

Properly managing the cultivation of cotton is essential because it directly impacts the amount of cotton that is produced. The aim of this work is the proposal of a fuzzy classification system for diagnosis-prediction tasks of the cotton crop yield. We used a soft computing method to handle/describe experts’ knowledge. Seven input variables (attack level of the red boll weevil, attack level of the black boll weevil, crop stage, rainfall, fertilizer, pheromone traps and boll-weevil killing tube) were considered in the system to analyze the cotton production. System tests were carried out on different agricultural scenarios, to determine their robustness and adaptability. According to the results, the fuzzy system has the capability to generate outputs that correspond with the experts’ evaluations, which can be used to help farmers select the best practices in cotton crop management, in order to obtain the best yield in a specific context. The developed models enhance our capacity to predict crop yields based on climate data, the soil and pest behaviors, a valuable indicator for decision-making and overall sustainability.