Prediction of the Population of Kapuas Hulu District Based on Gender Using the Backpropagation Method

Rediction is a branch of science used to estimate future events based on historical data. One of the effective methods currently developing is the Backpropagation Artificial Neural Network. This study aims to determine prediction results, the developed model, and its accuracy in forecasting the population of Kapuas Hulu district by gender using the Backpropagation method. The resulting model has an architecture of 2-5-2, with 2 neurons in the input layer, 5 in the hidden layer, and 2 in the output layer. The model uses a learning rate of 0.8, an error tolerance of 0.00001, and 8000 epochs. Predictions for one year after the last dataset year (2024) estimated 138,756 males and 131,434 females, achieving an accuracy of 99.38%. Model validation using the k-fold cross-validation method with 4-folds showed the best accuracy of 99.38% in the first fold. This indicates that the Backpropagation model is highly reliable and effective for predicting population data based on gender.