Development of a ForEx TradeBot.

The goal of this project is to develop a trading bot for ForEx market using modern Machine Learning techniques, such as Recurrent Neural Networks and Double Deep Q-learning Networks, as described in the paper "Deep Reinforcement Learning with Double Q-learning" by Hado van Hasselt (https://arxiv.org/abs/1509.06461).

My research is split into five parts:

1. Part 1 - Data exploration and preparation;
2. Part 2 - Building an Environment;
3. Part 3 - Building an Agent;
4. Part 4 - Designing a visual Testsight;
5. Part 5 - Training and Testing;

IPython notebooks contain in-depth explanations to every step of this journey, but are not necessarily RunAll'able, because they're quite old and I didn't bother updating them. Good news is that you can find working code in classes folder.

So feel free to explore. Unfortunately, you won't find data to reproduce this research for yourself, as I didn't bother uploading it yet :)