Our assignments for this course were very open-ended. We could create whatever we wanted as long as it addressed a topic in AI. The following two projects were completed with Henry Herzfeld (GitHub) (LinkedIn) and Yuri Villanueva:

One of the projects involved a deep reinforcement learning implementation of the board game Connect4. It is featured on the Deep Connect4 page in the "Home" section. I mention it here because it is also featured on Dr. Marques' blog, found here.

The second project involved an examination of the problem of Spam Detection from the perspective of each of the five tribes of machine learning described by Pedro Domingos in his book, The Master Algorithm. A notebook with a more detailed explanation can be found here. The work was again also featured in Dr. Marques' blog, found here.

Finally, there was also an individual project done by me that examined the effectiveness of a CNN on a Pokemon image dataset from Kaggle. A detailed notebook can be found here.