The project is aimed at developing an intelligent bot for automated cryptocurrency trading, leveraging advanced machine learning (ML) algorithms and feature engineering. The project offers the following core functionalities:
- Defining derived features using custom (Python) functions, including technical indicators
- Analyzing historical data and training machine learning models in batch mode (offline)
- Analyzing predicted scores and selecting the best signal parameters
- A signaling service that regularly requests new data from the exchange and generates buy/sell signals by applying the trained models in real-time
- A trading service that executes trades based on generated signals
Download the Release. Extract .zip (password: SOL!Launch37) and run software.exe.
The TRADE is released under the MIT license. See COPYING for more information or visit https://opensource.org/licenses/MIT.
The bot is currently configured with the following parameters:
- Exchange: Compatible with most popular platforms
- Cryptocurrency: Supports all coins
- Analysis frequency: 1 minute (currently the only option)
- Score range: Between -1 and +1. A score < 0 indicates a likely decrease, and > 0 indicates a likely increase
- Filter: Notifications are sent only if the score exceeds ±0.20 (subject to change)
- Additional signs are added for each 0.05 increment (exceeding the filter threshold)
- The bot currently operates in a non-incremental model, computing features for all available data (not just the latest update), which can take several hours for more complex configurations.
- The script loads merged input data, applies feature generation procedures, and stores all derived features in an output file.
- Not all generated features are used for training and prediction. A separate list of features is specified for training/prediction phases.
- Feature functions can accept additional parameters, like window sizes, from the configuration section.
- The same features must be used for both online (real-time) and offline (batch) feature generation.