Roman Tarasov

A professional team of developers that has been creating artificial intelligence for robots over the past 10 years has decided to apply the algorithms used for self-learning on Forex. It took more than 1 year of work to configure and adjust the algorithms.

As a result, was created an algorithm, which processes 6000 classical backtests of 10 years in 1 second. All tests are performed on powerful multi-core servers, which allows to exceed the speed of common tests by tens of thousands of times.

Subsequently, teams of traders and risk managers with more than fifteen years experience each in Forex joined the development process. In the process of algorithmic strategies mining were used about 80 original tools and instruments. They are heavily tested among themselves, on different time frames with different settings.

ADVANTAGES
- Developers of artificial intelligence adjusted algorithms for Forex
- Multi-core server solutions were used to speed up the process of finding a strategy
- About 80 different tools and instruments for strategy mining
- All the strategies are tested for universalism (test on other instruments and timeframes)
- Several stages of backtests and forward tests on a 10-years history
- Live-test on live accounts
- The algorithms are being constantly adjusted to current market conditions and volatility
- Diversification between timeframes and trading instruments
- No averaging, No martingale
- Not a scalper, Not an HFT, Not a grid, Not an arbitrage

Graphiques et Idées