In this project, a simple deep learning model is trainned, optimized, and used to predict the whether a team will win a match playing in its own stadium.
Results show a acuracy of 56.34% compared to a natural class distribution of roughly 50%.
Coloque aqui as ferramentas que você usou para criar seu projeto, exemplo:
The project implementation is still very simple. Next steps will enrich the project.
- Test more ANN architectures and other type of models
- Get more data to make models more powerful
- Train models to also predict multiclass target: win, withdraw, and lose
- If you want to contribuite or have any suggestions, reach me out on linkedin =)