Review of precious metal exchange rates forecasts

Precious metals have always played a fundamental role in financial activities, being treated as a global currency, an attractive investment and a raw material for many industries. Effective forecasting of the trend in precious metals has long been of interest to researchers in many fields, and such forecasts can be very useful in regulating the risk of personal investments and in adjusting the national economy in times of strategic change. Nowadays, mainstream forecasting methods include machine learning methods, among which a number of regression models and deep learning algorithms are widely used and accepted. Recent developments in deep learning offer opportunities to extract valuable information and make forecasts using the vast financial and socio-economic data available. One of the biggest advantages of these methods is that they can learn significantly more complex relationships between input and output data, even those that are difficult to define or even unknown at first, so they can effectively help professionals make their predictions. Therefore, the main aim of this paper is to summarize and compare the results of the recent forecasting using different deep learning algorithms.