Rainfall prediction in Tengger region Indonesia using Tsukamoto fuzzy inference system

Rainfall Prediction in Indonesia is very important for agricultural sector. However, obtaining an accurate prediction is difficult as there are too many input parameters including the world climate change that affect the accuracy. An accurate prediction is required to arrange a good schedule for planting agricultural commodities. A good approach is required to obtain a good model as well as the accurate prediction. This paper proposes Tsukamoto fuzzy inference system (FIS) to solve the problem. An intensive effort is put in building fuzzy membership function based on rainfall data in Tengger region from ten years ago. A series of numerical experiments prove that the proposed approach produces better results comparable to those achieved by other approach. In Tutur region Tsukamoto fuzzy inference system obtain Root Mean Square Error (RMSE) of 8.64, it is better than GSTAR-SUR method that obtain RMSE of 10.89.