Neural Network and Genetic Algorithm Approaches for Forecasting Bangladeshi Monsoon Rainfall

True information about rainfall is crucial for human activities such as the use and the management of water resources, hydroelectric power projects, warning to impend droughts or floods, urban areas sewer systems and many others. This paper investigates the development of an efficient model to forecast monthly monsoon rainfall for a number of stations, namely Barishal, Chittagong, Dhaka, Khulna, Rajshahi and Sylhet. It is believed that rainfall forecasting is difficult and also a challenging task for anyone because rainfall data are multi-dimensional and nonlinear. Therefore, to model rainfall data, the AI models, namely artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS) and genetic algorithm (CA) have been used. Results found by the AI models are also compared to the linear regression model to show advantages of selecting these models. Findings suggest that ANFIS and CA approaches could be used more accurately than the other two selected approaches to forecast the Bangladeshi monsoon rainfall.