Data Scientists Apply AI to College Basketball Tournament Bracket Predictions, But Can't Beat Unpredictability of March Madness
• Many U.S. sports fans enjoy filling out "March Madness" college basketball tournament brackets to predict winning teams.
• Some researchers are using artificial intelligence (AI) and machine learning to try to create the best mathematical models for predicting tournament outcomes.
• Creating a perfect bracket is extremely unlikely, even with advanced AI tools, due to the unpredictability of human competition.
• The website Kaggle hosts an annual "Machine Learning Madness" bracket competition for data scientists to test their algorithms.
• Experts say a balance of mathematical modeling and human intuition is needed to account for the randomness inherent in sports.