Prediction Of The Gold Futures Price Based On Deep Learning

Gold futures,as an investment product in the market,is more and more important in the view of hedging or the maintenance of financial stability.So the prediction of its price has some practical significance.At present,the study of its price mainly focuses on the factors that affect its change,the linkage with the price of other financial instruments and the prediction based on different models.It should be pointed out that the change of gold futures price is a very complex nonlinear dynamic system.If the forecasting method based on the traditional econometric models is used,there are some difficulties.And the shallow neural network is used to predict futures prices,there are many restrictions and shortcomings,too.However,deep learning,as a new field of study in machine learning in recent years,can portray various highly non-linear changes,and achieved good performance in quantitative investment,risk control,and promoted the application of artificial intelligence in the area of finance greatly.Therefore,this thesis attempts to use LSTM and GRU,which are two kinds of deep learning model that have the "memory" of the previous information,to predict COMEX gold futures prices.