Recent Research Advances in Machine Learning and AI

Graphs and networks are ubiquitous in both human-made and natural processes. The challenge is in understanding their behaviour and in optimising their performance. Computationally tractable probabilistic methods can help us address these challenges. I will review some of my recent research activities in the areas of improving internet throughput, testing and mitigation in epidemic spreading, the development of cortical neurons-based integrated circuits and better understanding of deep-learning machines.