A simple trading simulator inspired by the OpenAI framework. The objective of the project is to explore the viability of AI (Supervised/ Reinforcement) algorithms in the financial markets. Models built in Tensorflow. Designed to carry out historical backtesting and live trading with Oanda API.
- Base Environment Representation
- Experiment #1: Trading with previous n periods of Open-Open Return (DQN)
- Experiment #2: Trading with Moon Coordinates? (DQN)
- Experiment #3: Trading with A3C
Sources: DQN made simple by Arthur Juliani
- numpy
- tensorflow
- talib
- pandas
- swisseph (Swiss Ephemeris)
1.Installation:
pip install -e .
-
Enter Oanda API ID and token key in serverconfig.py
-
Best viewed in notebooks
This is by Henry Bee, intended for Mr Peter's use