This is an MLflow integrated pipeline built using Ploomber framework. Upon providing the required list of parameters, the input data will be processed and fed into the XGBoost Classifier which will be trained to predict in binary sense whether a given set of financials is neutral or bullish/bearish (e.g., it has reached the local maxima or minima).
The code itself can download the requested Crypto through using binance client (of which you'll need an API key provided in a keys.yaml file). This project in on development phase. Hopefully I will be fixing some bugs and adding the multiclass version of the work.
DISCLAIMER: This algorithm is built for learning purposes and it's NOT advised to use it on live trading
You can access the very first notebook output of the pipeline here: Notebook: report-0
Please refer to the img folder for the artifact files.