This article presents a method for modelling the onset and end dates of the monsoon season in Northern Madagascar. The method relies on an Artificial Neural Fuzzy Inference System (ANFIS) based on artificial intelligence, which utilizes daily zonal wind data at 925 hPa and a moving average smoothing technique to estimate the onset and end dates of the monsoon. The average period of the monsoon season in this region ranges from December 29th to March 5th. The onset and end dates of the monsoon season serve as inputs to the ANFIS model, which consists of four inputs and one output. To improve forecast accuracy, forecast accuracy, the model results are compared to observation data and validated using the root mean square error (RMSE) criterion.