A Comparative Study of Mamdani & Takagi-Sugeno-Kang Fuzzy Inference Systems for Rainfall Intensity Modeling in Mataram City

Authors : Nurul Ulya Ayudia; Mamika Ujianita Romdhini; Syamsul Bahri
article cite 0 Year 2026
source: IOP Conference Series Earth and Environmental Science
Abstract

Abstract Weather is the atmospheric condition in a certain region with limited geographical coverage and occurs within a relatively short period of time, related to the state of the air on Earth. In certain conditions, weather forecasts may be inaccurate or inconsistent with actual conditions. This is because rainfall does not always follow a consistent pattern. To address this issue, this study aims to apply fuzzy inference system (FIS) models of the Mamdani and Takagi-Sugeno-Kang (TSK) types to predict rainfall levels in Mataram City and compare their accuracy. The data used are daily weather records from the Meteorology, Climatology, and Geophysics Agency (BMKG) for the period of January to March 2025, including temperature, humidity, sunlight duration, wind speed, and rainfall on the previous day. The results show that the TSK FIS model outperforms the Mamdani FIS model in both modeling and testing datasets. In the modeling data, the RMSE values for the TSK and Mamdani FIS models are 16.61 and 18.40, respectively, while in the testing data, the RMSE values are 20.28 and 22.60, respectively. Moreover, the linguistic accuracy of the TSK FIS model is also superior to that of the Mamdani FIS model. Therefore, the TSK FIS model is considered more accurate in modeling rainfall prediction in Mataram City.


Concepts :
Hydrological Forecasting Using AI
Water and Land Management
Multimedia Learning Systems
article cite 0 Year 2026 source IOP Conference Series Earth and Environmental Science
SDGs
Climate action
Citations by Year
YearCount
2026 0