<![CDATA[This study aims to predict rainfall in Muaro Jambi Regency using the Backpropagation Artificial Neural Network (ANN) method. The input variables include air humidity, air temperature, air pressure, and wind speed, with data obtained from the BMKG Muaro Jambi Climatology Station. The method is quantitative with a time series approach, involving data collection, normalization, and division into training, validation, and testing, along with the application of Trainlm, Trainrp, and Traindx. The results show that air humidity has the greatest influence on rainfall, while temperature, air pressure, and wind speed show weak negative correlations. Testing variations in the number of neurons in the hidden layer shows that 100 neurons with the Traindx algorithm produce the best performance, with a Mean Square Error (MSE) of 4.95%, categorized as very accurate. The Backpropagation ANN model follows the actual rainfall pattern from BMKG with a conformity level of more than 95% and recognizes seasonal patterns such as peak rainfall in March and a decrease in the middle of the year. Thus, this model is effective for predicting rainfall and supports disaster mitigation planning and water resource management in Muaro Jambi Regency.]]>
Copyrights © 2026
