Machine Learning Seminar: Beyond NIPS/ICML Publications
In this talk I will discuss both learning fundamental results and real applications and challenges. The theoretical foundations provide some insights and potential directions for improvement of common learning successes, and mostly follow the ideas of geometric regularization. The applications range from successes (face analysis) to challenges (mostly in developmental and mental health). The discussion will hopefully provide enough motivation for students not just to target publications but to target solving real problems.