A positive EV framework for NFL betting using Deep Learning
A back-tested framework here was used to make intelligent decisions based on a RNN model trained on historical betting/NFL data.
Author Archives: aannestrand
Chess Vision
Chess Vision
In this project, I created a sequence of algorithms to interpret and translate a picture of a chess board to digital form.
On predicting offense formation pre-snap
On predicting offense formation pre-snap
In this project, a small dataset of pre-snap images were collected for image processing and machine learning. The goal of the project was to explore the possibilities of ML in the realm of football analysis. Traditional image processing techniques such as Hough Transform and image gradients were utilized as pre-processing techniques for a CNN classifier. See details and all project report here:
Test Post
Test Post
This is a test blog post.
Andrew T. Annestrand 