Comparative Analysis of ARIMA and LSTM Algorithms for Predicting Chilli Prices in Bali
Ifan Prihandi, Sutarto Wijono, Irwan Sembiring, Evi Maria · 2024
This study evaluates the performance of two predictive algorithms, Autoregressive Integrated Moving Average (ARIMA) and Long Short Term Memory (LSTM), in forecasting the prices of red bird’s eye chilli in Bali Province, a crucial task for market stability. The objective is to identify the more accurate model using Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) as evaluation metrics. Historical price data were analyzed, revealing that the ARIMA model demonstrates superior accuracy with an MAE of 12068.53 and a MAPE of 31.37%, compared to the LSTM model, which has an MAE of 12478.11 and a MAPE of 33.63%. These results indicate that ARIMA is more effective for predicting chilli prices, as evidenced by its lower error values..