A DEEP LEARNING APPROACH FOR FORECASTING GLOBAL COMMODITIES PRICES

Forecasting future values of time-series data is a critical task in many disciplines including financial planning and decision-making. Researchers practitioners statistics apply traditional statistical methods (such as ARMA, ARIMA, ES, GARCH) for long time with varying. accuracies. Deep learning provides more sophisticated non-linear approximation that supersede most cases. require minimal features engineering compared to other methods; it adopts an end-to-end methodology. In addition, can handle huge amount variables. Financial series forecasting poses challenge due its high volatility non-stationarity nature. This work presents hybrid deep model based on recurrent neural network Autoencoders techniques forecast commodity materials' global prices. Results showbetter accuracy regression short-term horizons (1,2,3 7days).