Publication

Abstract : Crude Oil is one of the world's most essential commodities, and its price greatly influences the global economy. Both the public and corporate sectors place high importance on accurate crude oil price forecasts because they can influence decision-making and lower economic risk. Artificial Neural Net-works(ANN) and other deep learning architectures like Long Short Term Memory(LSTM) and Gated Recurrent Unit(GRU) have been employed in a variety of ways in the past to predict crude oil prices. In this paper, the LSTM-GRU hybrid method is proposed. The obtained result is compared with LSTM and GRU models. The forecasting has been done using each day's prices of Crude Oil from July 01, 2002, to July 01, 2022. The challenges faced while predicting are long-term dependencies, overfitting, and hyperparameter tuning. Mean Absolute Error(MAE), R squared error and Mean Squared Error (MSE) were used to assess how well the models performed when making predictions.