Munich Personal RePEc Archive

Generative Adversarial Network for Market Hourly Discrimination

Grilli, Luca and Santoro, Domenico (2020): Generative Adversarial Network for Market Hourly Discrimination.


Download (803kB) | Preview


In this paper, we consider 2 types of instruments traded on the markets, stocks and cryptocurrencies. In particular, stocks are traded in a market subject to opening hours, while cryptocurrencies are traded in a 24-hour market. What we want to demonstrate through the use of a particular type of generative neural network is that the instruments of the non-timetable market have a different amount of information, and are therefore more suitable for forecasting. In particular, through the use of real data we will demonstrate how there are also stocks subject to the same rules as cryptocurrencies.

MPRA is a RePEc service hosted by
the Munich University Library in Germany.