Gold Price Forecasting Using LSTM, Bi-LSTM and GRU

Due to the multifactorial and non-linear nature of gold market, it is difficult predict price. The price affected by many external factors, such as market environment, economic crises, oil increases, tax advantages interest rates. Therefore, multivariate models can better than univariate models. This study investigated effects price, crude exchange rate index, stock indicators between 2001 2021. Models created using LSTM, Bi-LSTM GRU methods were evaluated lowest Root Mean Square Error (RMSE), Absolute Percent (MAPE) (MAE) metrics. LSTM model performed best, with 3.48 MAPE, 61,728 RMSE 48.85 MAE values.