NHL Data Science Project
Type: IFT 6758 (Data Science) Project
Year: 2021
Teammates: Jack White, Sal Elkafrawy
Key Skills: Python; Data Science Pipeline; Data Science Libraries; Machine Learning Libraries; Docker
Additional Information: Project Repository
Objectives:
- Practice data science and machine learning skills by analyzing and visualizing the NHL data and developing and deploying a predictive model.
Summary:
The project is separated into three phases. The first phase focuses on acquisition, cleaning, and visualizing the NHL data. The second phase focuses on feature engineering and developing a predictive model to determine whether a shot scores. The third phase focuses on model deployment. The work is divided evenly among the team.
- Retrieved the NHL data and accessed the deployed model through REST API.
- Organized and performed feature engineering using Panda.
- Visualized the data information using Matplotlib and Seaborn.
- Created interactive interfaces in Jupyter Notebook using Ipywidgets and Plotly.
- Created statistical models and machine learning models for goal prediction using Scikit-learn.
- Deployed the model as a Flask application on Docker.
</div>