Valuing Basketball Players with Neural Networks

I was recently approached about figuring out a better way for college basketball teams to rank prospects given their high school player statistics.  I was eager to work on this project because I had been learning a lot about neural networks and thought they could provide a lot of insight into this problem.  After doing a lot of work with the data I concluded that an Excel plugin that could create neural networks for any data would be very convenient. 

The tool that I ended up creating turned out to be very easy to use and has the potential to put the power of neural networks into many more people’s hands.  The fact that it is an Excel plugin means that users can easily manipulate their data while running tests.  Also, the output and graphs are in Excel format, which are easy to use and most people have experience with them.

The software requires just a basic knowledge of neural networks, but can also be used more powerfully by experienced users.  In order to create the neural network you simply fill in the inputs on the menu and then highlight your data.  Next you will see a new sheet created with your neural network and all of the output.  The neural network is even interactive, you can change the inputs and see the predicted output.

This software will work with any data that is appropriate for neural networks.  It can do things like value sports players, predict time series data, and do image recognition.