PERAMALAN HARGA CRUDE OIL MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) DALAM RECURRENT NEURAL NETWORK (RNN)

Syahira Rahmadhani Siregar (Universitas Islam Negeri Sumatera Utara)
Rina Widyasari (Universitas Islam Negeri Sumatera Utara)

Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika
Vol. 4 No. 3 (2023): Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistik

Crude oil or petroleum is a very important requirement in meeting world energy consumption. Every country definitely needs a supply of petroleum to fulfill their needs. Fluctuations in oil prices are always considered as a barometer of the economy throughout the world, so any change in oil prices is always an interesting topic to be discussed in the economic environment in every country. Therefore it is necessary to predict the price of petroleum, while the method used to predict oil prices is the Long Short Term Memory method, this study aims to predict future crude oil prices based on historical data using the Long Short Term Memory method, knowing the accuracy Forecasting crude oil prices and increasing market efficiency to be more efficient in allocating resources and in this study resulted in an RMSE accuracy of 2,665 and 2.7% Mape for data starting in 2018-2023, while for data for 2020-2023 it produces an RMSE accuracy of 2,630 and MAPE is 2.9%.