The goal was to create a model to get events which then a Random Forest Classifier can then be fit on certain statistical properties to see an improved accuracy.
GitHubThis project takes real network traffic data which is annotated between malicious and not malicious and feeds that data into a bidirectional RNN to achieve an AUC of 0.83
GitHubCreated a Convolutional Neural Network that attempts to predict company stock price increase/decrease.
GitHubAssisted a team attempting to recreate AtomNet, a neural network that can predict the binding of proteins and ligands.
GitHubRecreated a small part of AlphaGo to predict the outcome of a match of a board state of Go.
GitHubCreated a reinforcement learning AI capable of simple squad based combat play in StarCraft:Brood War.
GitHub