MODEL JARINGAN SARAF TIRUAN UNTUK PREDIKSI PERMINTAAN PRODUK UMKM DI PEMATANG SIANTAR

This study aims to develop an Artificial Neural Network (ANN) model in predicting demand for MSME products in Pematangsiantar to optimize production and inventory management. The main problem faced by MSME actors is demand uncertainty which causes excess or shortage of stock, thus affecting business efficiency. The ANN model is applied with a guided learning approach using the backpropagation algorithm to analyze demand patterns based on historical sales data. Data were obtained from the Cooperatives and MSMEs Office of Pematangsiantar City and interviews with business actors. The research process includes data collection and pre-processing, variable selection, data sharing, model development, training, optimization, and evaluation using the Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percent Error (MAPE) metrics. The results of the study show that the ANN model with the backpropagation algorithm is able to provide accurate demand predictions, with a MAPE value below 10%, which indicates very good forecasting. The implementation of this model helps make it easier for MSMEs to make strategic decisions related to production and inventory, thereby increasing competitiveness in the market.