Application of Fuzzy Inference System Takagi-Sugeno Methods for Diagnosis of Diseases and Pests of Madura Paddy Rice Based on Symptoms

Abstract

Diseases and pests of lowland rice are one of the factors that can cause a decrease in rice yields. Therefore, it is necessary to have a diagnostic system to identify diseases and pests of paddy rice from an early age based on damage symptoms. The process of diagnosis requires expertise, knowledge, and experience from experts. Therefore, this research tries to build an expert system that can diagnose diseases and pests of paddy rice early by applying the Fuzzy Inference System Takagi with the Sugeno method. Fuzzy Inference System Takagi forms fuzzy sets using implication functions (rules). Rule composition is obtained from a data set of relationships between regulations, where the affirmation (defuzzification) and input from defuzzification is a constant or linear equation. The Sugeno method is used to diagnose diseases and pests of rice plants based on the symptoms experienced. This research aims to help plant pest control officers diagnose diseases and pests of paddy rice plants from the symptoms that attack the rice. The testing technique used is system accuracy testing and Mean Opinion Score (MOS) testing. The MOS test was carried out by involving 30 respondents consisting of 10 farmers and 20 extension workers, where 4.27 was obtained on a scale of 5 which was categorized into a good system. while testing the accuracy obtained from testing the system on two experts on diseases and pests of Madura paddy rice plants in 30 different cases has resulted in an accuracy rate of 86.66%. The expert system built in this study was able to diagnose 13 diseases and pests of Madura paddy based on the knowledge of two experts on 38 symptoms, and the plan was feasible to use and categorized into a good system.