Analysis of the LSTM Model for Rice Price Prediction at Makmur Abadi Store in Bengkulu

Abstract

Rice is a key commodity whose price fluctuations directly affect community welfare and market stability. Accurate price prediction can help traders, consumers, and policymakers make better economic decisions. This study developed a web-based rice price prediction application using the Long Short-Term Memory (LSTM) method. Historical rice price data from Toko Makmur Abadi were processed using Pandas and NumPy for data manipulation, normalized with Scikit-learn’s MinMaxScaler, and modeled using Keras with TensorFlow as the backend. The web interface was built with Streamlit, while Plotly was used for dynamic visualization of prediction results. The LSTM model was trained to predict rice prices for early 2024 based on time series data. The results show that the system can operate effectively, providing accurate and easily interpretable predictions through an interactive web interface.