Trading System on the Foreign Exchange Market. Evolutionary Reinforcement Learning Approach

Exhibitor: Andrei Hryshko

Supervisor: Tom Downs

Research Group: Complex and Intelligent Systems

Industry Sector: Finance

Foreign Exchange trading has emerged in recent times as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process will be very helpful. Here we try to create such a system with a Genetic Algorithm engine to emulate trader behaviour on the Foreign Exchange market and to find the most profitable trading strategy.

The main instrument employed by traders is the set of available indicators which helps traders to discover trends, trend reversals and other fluctuations. Usually traders start with some concepts based upon indicators and then turn those concepts into a set of objective rules. The rule creating process requires the trader to choose which indicators to rely on and to use his own subjective opinion and intuition to define the rules on how to interpret the indicator signals.

Described here is the design of a system engine based on machine learning and embed it into a trading system. This system will draw upon available information to determine the optimum strategy for the trader. Moreover, unlike the human trader, it will work on-line and so will update its parameters over time to achieve the highest returns. This system will take decisions and predict the future market based on a combination of different market models. The system will automatically recognise the condition of the market based on the simultaneous examination of signals from all indicators (instead of examination of indicator signals one by one).