FEUP M.EIC Y1S1
To run the project, check out our documentation provided in docs/source/index.rst. Access html pages under that directory to see all the needed knowledge in a formatted web page.
We are a team of students trying to predict WBNA basketball results. WBNA (Woman's National Basketball Association) is a domestic basketball league hosted in the United States that features 12 different teams. Each season, 8 teams qualify for the playoffs - the final round of matchups to determine the champion.
The main objective of this project is accurately predict which teams will qualify for the WNBA playoffs in the next season based on the provided data.
This data has been collected through the past decade, and it congregates information on the players, teams, coaches and winning-awards for each season.
We'd hope to consistently predict atleast 6 out of 8 teams that reach the playoffs, resulting in a optimistic success rate of
The team is made of 3 students from the M.EIC degree in University of Porto. Each possesses their own laptop to further research on the topics related to the problem. There is a dataset with information on the last 10 WNBA seasons.
The model must be ready within 12 weeks and the team is aiming for a success rate of 75 percent, as mentioned before. The team is limited by the computing power offered by Google Services (like Google Colab) and their own hardware.
Worst case scenario, the project might be delayed in the event of the University scheduling multiple exams near the date of submission.
The following terms will be used in the context of this project:
| Terminology | Meaning |
|---|---|
| NN | Neural Network |
| TF | Tensor Flow |
| MLP | Multi-Layer Perceptron |
The project has 0 monetary costs associated, as its being done for free, for educational purposes. That means profit is surely guaranteed. From a business standpoint, one could use this model to gamble their way into accurately predicting next year's playoffs bracket - generating money, most probably.