TraderNet-CRv2 - Combining Deep Reinforcement Learning with Technical Analysis and Trend Monitoring on Cryptocurrency Markets
This system architecture is an extended version of the original TraderNet-CR architecture, which is described by this paper: https://link.springer.com/chapter/10.1007/978-3-031-08333-4_25. In this work, we combine Proximal Policy Optimization algorithm (PPO), which is a DRL learning algorithm, with 2 rule-based safety mechanisms: N-Consecutive & Smurfing. Our experiments on 5 popular cryptocurrencies show very promising results.
Technical analysis has been applied on market data in order to train TraderNet. The following popular technical indicators have been used:
- EMA (Exponential Moving Average)
- DEMA (Double-Exponential Moving Average)
- MACD (Moving Average Convergence/Divergenc)
- AROON
- CCI (Commodity Channel Inde)
- ADX (Average Directional Inde)
- STOCH (Stochastic Oscillator)
- RSI (Relative Strength Index)
- OBV (On-Balance Volume)
- BBANDS (Bolliger Bands)
- VWAP (Volume-Weighted Average Pric)
- ADL (Accumulation/Distribution Line)
To run and evaluate our agent, You need to download the following libraries/packages:
- Python
$\geq 3.6$ (https://www.python.org/) - Numpy (https://numpy.org/)
- Pandas (https://pandas.pydata.org/)
- Matplotlib (https://matplotlib.org/)
- Tensorflow (https://www.tensorflow.org/)
- TF-Agents (https://www.tensorflow.org/agents)
- TA (https://technical-analysis-library-in-python.readthedocs.io/en/latest/)
- PyTrends (https://pypi.org/project/pytrends/)
- Scikit-Learn (https://scikit-learn.org/stable/)
Download Python 3.6 or higher and the libraries that are described on requirements using pip installer (e.g. pip install numpy). Then:
- Run
download_datasets.pyto download the datasets from CoinAPI platform (https://www.coinapi.io/). - Use
train_tradernet.ipynbto train TraderNet module. - Use
train_smurf.ipynbto train Smurf module. - Use
integrated.ipynbto evaluate the Integrated agent.
- Bitcoin (BTC)
- Ethereum (ETH)
- Cardano (ADA)
- Litecoin (LTC)
- XRP
This AI is not a commercial product and is intended for research purposes only.