PERBANDINGAN MODEL LSTM DAN GRU UNTUK MEMPREDIKSI HARGA MINYAK GORENG DI INDONESIA

Abstract

Cooking oil, a food ingredient used for cooking, has increased in price in Indonesia. Based on the Indonesian Strategic Food Price Center data, cooking oil reached twice the regular price at the beginning of 2022. Long Short-Term Memory and Gated Recurrent Units are several ways to predict time domain data. The data set for prediction is cooking oil price data for the last three years in the Indonesian Strategic Food Price Center data. The data taken is pre-processed by changing the data type and filling in the blank day data with the previous day's value. Data that has been pre-processed is modeled using Long Short-Term Memory and Gated Recurrent Units. Modeling obtained Error, Loss, and selection of the best unit and dropout model parameters in making predictions for both types of models. As a result, this research compared the prediction results of the Long Short-Term Memory model and the Gated Recurrent Unit on the cooking oil data set and predict price changes that will occur during the following year.

Keywords

• Cooking oil
• Gated Recurrent Unit
• Long Short-Term Memory