icbc_practice

Simple financial instrument prediction solution using ML regression models.

Prerequisites

• Tested on Python 3.11
• Virtual environment, see https://github.com/pyenv/pyenv
• pip intstall -r requirements.txt
• Datasets are not commited, add them and define the paths in config.py
• See config.py and adjust any other confiuration as required

Run locally

python main.py

Run in Docker

docker build . --tag icbc

docker run -v ${pwd}/output:/usr/src/app/output icbc

Additional notes

• Depending on your operating system, you might require to manually create the output directory on your host when running in Docker.
• GridSearchCV is time-consuming, you might want to just predict the target, once you have trained a model. Use the Regressor.predict_with_model method