Event-driven approach for stock return forecast

Project Description

This project uses Machine Learning to analysis the impacts of various events to company's stock price and use it to predict future price changes.

34 different types of pre-classified events, from Departure of Directors to Failure to Make a Required Distribution, have been considered.

Results

Confusion metrics:

TP: 637 FP: 183
FN: 333 TN: 142

Usage

To run the model, run python3 main.py

Development

Prepare dataset in datasets.py Prepare the corresponding model in models.py

Built With

Share this project:
×

Updates