An RNN model to predict football player performance given a sequence of five past match statistics.
-
WebApp: This directory contains everything needed to run the webapp for the model locally. In order for the webapp to compile, the following Python modules must be installed locally or on a virtual environment: BeautifulSoup, pandas, requests, Pytorch, numpy, Flask, lxml.
-
data: This directory contains all of the code used to gather the data used to trained the model.
-
model: This directory contains the code to train the model and the weights produced by it, but not the dataset used to train or validate it (the training set was too large to upload on github).
-
Deliverables: This directory contains the deliverable reports for MAIS 202.
-
code: This directory contains mostly code for previous, discarded iterations of the model.