Joint Workshop KO2PI and The 1st International Conference on Advance & Scientific Innovation

Research Article

Rainfall Forecasting Using Backpropagation Neural Network: A Case in of North Luwu, Indonesia

Download471 downloads
  • @INPROCEEDINGS{10.4108/eai.23-4-2018.2277580,
        author={Ansari Saleh Ahmar and Muhammad Arif Tiro and Ria Putri Utama},
        title={Rainfall Forecasting Using Backpropagation Neural Network: A Case in of North Luwu, Indonesia},
        proceedings={Joint Workshop KO2PI and The 1st International Conference on Advance \& Scientific Innovation},
        publisher={EAI},
        proceedings_a={ICASI},
        year={2018},
        month={7},
        keywords={rainfall forecasting},
        doi={10.4108/eai.23-4-2018.2277580}
    }
    
  • Ansari Saleh Ahmar
    Muhammad Arif Tiro
    Ria Putri Utama
    Year: 2018
    Rainfall Forecasting Using Backpropagation Neural Network: A Case in of North Luwu, Indonesia
    ICASI
    EAI
    DOI: 10.4108/eai.23-4-2018.2277580
Ansari Saleh Ahmar1,*, Muhammad Arif Tiro1, Ria Putri Utama1
  • 1: Department of Statistics, Universitas Negeri Makassar, Daeng Tata Street, Makassar, Indonesia
*Contact email: ansarisaleh@unm.ac.id

Abstract

Climate information is seen as important in helping society, especially in the agricultural sector. In this regard, it is necessary to forecast rainfall which is one of the major factors associated with global climate change. ARIMA Method is a popular method used in forecasting. However, the method in use was based on several assumptions and is not considered independent variables in making forecasting. To anticipate the non-fulfillment of assumptions in the ARIMA method Box - Jenkins and to further streamline the results of forecasting by using free variables in forecasting, another method that can be used is backpropagation neural network. This study aims to predict rainfall in North Luwu District use backpropagation neural network method. From the research results obtained the best model i.e. architecture (5,4,1) for backpropagation neural network method with RMSE 95,41.