Skip to content

Data Analysis and Model Building for 2020 Kaggle/Google Cloud NCAA March Madness Tournament.

License

Notifications You must be signed in to change notification settings

asoane34/NCAA_March_Madness_2020

Repository files navigation

NCAA_March_Madness_2020

Data Analysis and Model Building for 2020 Kaggle/Google Cloud NCAA March Madness Tournament.

Repository Contents:

00DataSetCreation-FeatureEngineering

Data wrangling and manipulation. Extracting data from .csv tables provided via Google Cloud and and engineering team statistics (both game level regular season and postseason) to prepare a feature set for modeling.

01.1UpsetPrediction

Data analysis for trends and patterns in predicting upsets. Designed model to predict upsets (potentially used in final voting ensemble model)

01 EDA-TournamentModeling2003-2019

EDA and model building using only end of regular season data available from 2003 - 2019. Primarily used as a benchmark for final model.

02.1RegularSeason-StatisticsOnly

Model building using only team statistics from regular season games to predict outcomes. Level 0 model.

02.2RegularSeason-RankingsOnly

Model building using only rankings and ratings from regular season to predict outcomes. Level 0 model.

02RegularSeasonModel

Model using both regular season statistics and rankings. Benchmark for ensemble method.

03ModelEvaluation

NOT COMPLETE- evaluating effectiveness of models and experimenting with model combinations.

About

Data Analysis and Model Building for 2020 Kaggle/Google Cloud NCAA March Madness Tournament.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages