For avid sports fans, it can be difficult to find information about statistics and probabilities for your favorite teams without paying or wading through piles of non-information. We want to provide a service that uses well-respected data sources and models to aggregate statistics and provide sports predictions for free.
- Build a robust web application which displays accurate predictions for winners of upcoming sports games.
- Provide the ability to create an account on the above website and track selected teams.
- Develop a custom machine learning algorithm to predict game outcomes based on pre-game data.
- Store team and prediction data in a database to reduce API calls.
- Host a fully functioning website that is accessible from the web.
- Track games in real time and get updated predictions and statistics while the games are progressing.
- Provide ancillary information about the sporting events and the participating teams.
- JavaScript
- Python
- Node.js
- Express.js
- React.js
- MongoDB
- Google Cloud
- Clone the Repo
- Run
./install.sh- This will install all dependencies for both the frontend and the backend
- Run
pip install -r requirements.txtin/alderaan,/mustafar, and/hoth- This will install all dependencies for the python services
- Run
npm run startin the root to start the frontend and backend - Navigate to
localhost:3000
Note: Code additionally requires a config file in the following format to be placed in /backend
{
"databaseURI" : "",
"secret" : "",
"email_service": "",
"email_user": "",
"email_pass": "",
"nbaApiKey": "",
"TWILIO_ACCOUNT_SID": "",
"TWILIO_AUTH_TOKEN": "",
"TWILIO_PHONE_NUMBER": ""
}
The /frontend directory holds all files related to the running and deployment of the web application.
/src/componentsholds all custom React components for the frontend/src/assetsholds all static assets that are used on the frontend like team logos, and the DubbClub logo
The /backend directory holds all files related to backend web service which supports the website.
/src/servicescontains the majority of all logic for the website’s functionality, including all sports features, authentication, and connection with servers running ML processes./src/database/modelscontains our database schemas.
The /alderaan directory holds all files for EPL predictions (models, training data, and microservice).
The /hoth directory holds all files for MLB predictions (models, training data, and microservice).
The /mustafar directory holds all files for NBA predictions (models, training data, and microservice).
This project is Licensed under the GNU General Public License v3.0.