This project tests multiple different machine learning algorithms that can detect adversarial attacks in multi-agent reinforcement learning settings. Baselines were used to compare performance of a proposed ensemble model. Then, using FGSM, we re-attacked the ensemble detection model with perturbed observations. Read more at the pdf titled LischkeSeniorThesisFinal.
Of the machines tested, there were: -Knearest Neighbors -Random Forest -SVM -Binary Dense Neural Network -Binary LSTM Neural Network -Proposed stacked ensemble of Binary LSTM and predictive LSTMs