0 Project complete
We tried to tackle the problem of anomaly detection with a data set consisting of cyberattacks. As our baseline, we chose to apply the LightGBM algorithm. To improve the results of the baseline, we wanted to try auto-encoder, variational auto-encoder and generative adversarial network (GAN) methods. Our main challenge was the class imbalance problem in the data set. In other words, there were too many “normal” cases and there were only a few anomalies. Therefore, precision and area under the ROC curve were our primary metrics.
You can find the github page for this study group projects.
Topic;Teaching an AI to play games by using reinforcement learning algorithms which learn how to behave through a positive - negative reward system. We started with toy problems such as Taxi, Cart Pole and Pong games that have discrete action spaces on OpenAI’s gym environment. After successfully training AI programs to play these games, We moved onto continuous action space games: Mountain Car and Bipedal Walker. We tried different approaches but for the Bipedal Walker game, best results are achieved using genetic algorithms.
As a part of the course, we study the topics below by using Python language with the Associate Prof. Uzay Çetin.
Deep Learning Study Group II is a 16 week-long study group, in which we cover advanced deep learning study series for AI enthusiasts and computer engineers. We follow up materials on https://www.deeplearning.ai each week and get together on Saturdays to discuss them.