TripAdvisor Challenge, Brown Datathon 2017
This project is an exploration into data analysis using Matlab's neural network toolbox. Given a .csv file filled with TripAdvisor website usage statistics, our team of four attempted to predict whether a given user on that site would make a purchase before leaving.
Our program is able to predict these purchases with an overall accuracy of ~82.3%. There are four main files being used: getDayOfWeek.m, loadForNNData.m, loadSessions.m, and runNN.m. To run our neural net, just run the last two files: loadSessions.m and runNN.m.