The DATA SCIENCE REPORT has created a NEW section called PhD Predictive Projects! In this section we are going to profile interesting Predictive Analytics Projects by “up and coming” PhD and Masters students/recent grads. The exposure may land them that first great job or simply start a compelling dialogue.
All projects will be promoted here on the Data Science Report, externally to hiring managers, and relevant social media outlets, etc. We will decide together on how to market the work so it gets noticed. Here is very compelling project to start…
PROJECT #1 “BEATING THE SPORTSBOOK”
PhD Candidate (Expected Grad 2015)
Electrical / Biomedical Engineering
Using Machine Learning, Eric created a computer model in C# and Matlab which predicts the outcome of mixed martial-arts (MMA)/UFC fights and calculates how much to bet on each fight. It started as a for-fun side project – an experiment to see if a computer model could make predictions at least as accurately as human MMA/UFC fans. He didn’t really expect it to succeed, but it represented a technical challenge he found interesting.
The easiest place to test the model against humans is the betting market, as the betting odds represent the collective predictions of thousands of MMA enthusiasts. Using a hypothetical 10,000 bankroll to start, the model’s predictions between January and today have a total return of 160%! Professional gamblers would call that amazing!
So we’re now following the model’s predictions through the next several months, to see if it keeps up this level of performance. View Project and See Predictions Here.
PROJECT #2 “Your Project Here”
If you are a Masters or PhD Candidate/recent grad with a predictive analytics project you’d like us to profile, then please email me, Ted O’Brien, Directer Data Science & Advanced Analytics Practice & Publisher of the Data Science Report Email: [email protected]