NCAA March Madness is one of the most exciting sport events in the US and absolutely a bunch of folks want to bet on their favourite teams. TThe goal of the competition is to utilize past tournament data (from 2002 to 2019) to build and test predictive models in order to forecast outcomes of the Final Four in the 2020 NCAA Division I Men’s Basketball Championship. These outcomes are computed probabilistically, and the models are evaluated by log loss. For instance, Team 1 has 70% likelihood of winning Team 2. Also, if you are not familiar with log loss, just know that the best model has the lowest log loss. Even though the NCAA cancelled the tournament because of COVID-19 in 2020, I, and other participants still decided to try our hand at modeling the tournament and kept the tradition at Fordham going.
Navigate to notebook for a more write-up.
-
Notifications
You must be signed in to change notification settings - Fork 1
Engineered new features, performed exploratory visualization with Tableau, and applied Logistic Regression, Random Forest, Grid-Search and Cross Validation to build a predictive model that forecasts outcomes of the NCAA March Madness championship.
bigforehead/NCAA-March-Madness-Bracket-Prediction-2020
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
Engineered new features, performed exploratory visualization with Tableau, and applied Logistic Regression, Random Forest, Grid-Search and Cross Validation to build a predictive model that forecasts outcomes of the NCAA March Madness championship.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published