TensorFlow DevPost: Setup

While looking through all of the content from TensorFlow Dev Summit I saw that TensorFlow had created a DevPost challenge. As I have been talking with my friends about creating a machine learning project to help predict scores in college basketball I decided I would enter to see what would happen.

Setup: Since the Alpha release of TensorFlow was announced I decided to install that in a new environment and get started. I used my existing post about setting up my TensorFlow testing environment with the only change is that I used the Alpha build instead of the nightly build.

Data Collection: For the data collection part I decided I would scrape the NCAA web site of all the games from the 2018-2019 year.

Next: In a future post (hopefully next week) I will cover the creation and thoughts around the Juypter Notebook I created.

 

TensorFlow DevPost Submission

While looking through all of the content from TensorFlow Dev Summit I saw that TensorFlow had created a DevPost challenge. As I have been talking with my friends about creating a machine learning project to help predict scores in college basketball I decided I would enter to see what would happen.

Here is my submission: https://devpost.com/software/basketball-scout

In 2 future posts I will explain what went into building this project.

YouTube: Binary Classification

Video 1: Binary Classification.

In this video I walk through a side project I worked on for a friend. I took 8 games worth of football plays and tried to predict the next play. It was interesting to work from scratch on a project like this. In the end I was pretty happy with how it turned out. We will see next fall if it adds any real value though!