DPhil for Rasheed

Three cheers to Rasheed El-Bouri for passing his doctoral viva!  His research, published at ICML and others, focused on novel techniques for using reinforcement learning to for "student-teacher" learning with deep architectures, whereby the learning process for a model that one seeks to train (the student) is guided by a different model (the teacher); in this framework, the teacher creates a "curriculum" of examples through which the student will learn, starting with "easy" examples from the training set before moving on to "difficult" examples.  

His work focused on applications in improving patient flow within hospitals, which has become especially important during the COVID-19 pandemic, where Emergency Departments are under great pressure.