1. Problem Statement

2. Data extraction

3. Exploratory Data Analysis

4. Feature Engineering

5. Feature Selection

6. Model Selection

7. Model tuning

8. Acknowledgements

9. License

The purpose of this project is to produce a model to predict the winners of the 2023 March Madness college basketball tournament.

• All data will be acquired through Beautiful Soup from: https://www.sports-reference.com/cbb/
• The programming will be done inside of jupyter notebooks
• The data will be modified & analyzed using Python w/ the Pandas package
• This project will help to practice newly acquired data science skills

VSCode	Jupyter	Python	Github	iTerm2	MacOS
Code Editor	Notebook	Language	Developer Platform	Terminal	Operating System

Numbers	SciKit Learn	Feature Engine
Data Visualization	ML Library	Data Transformation

Data will be collected for all NCAA College basketball teams on Sports Reference website: https://www.sports-reference.com/cbb/

The data to be collected:

• Team Names (ex. Colorado Buffaloes)
• Team Home Page Link ex. 0,1,2)
• Team Years (years containing stats) (ex. '2002-03')
• Stat Labels (of data to-be collected) (ex. fg_pct)

Actions to be performed:

• Establishes 'Team' Object (https://github.com/wmauz677/MarchMadness2023/blob/main/Classes/team.py)
• Creates teams_dictionary: the main data dictionary where all stats will be stored for all teams for all years
• Writes teams_dictionary to 'Data/teams_dictionary.pkl'

Data Source

• Sports Reference
• Sports Reference Robots.txt

Distributed under the MIT License. See LICENSE for more information.