How DID YOU GET INTO machine learning?
I started my research career with an interest in computer graphics and differential geometry. During PhD studies in the UK I found myself gravitating towards computer vision, in particular 3D reconstruction with applications in biometrics. After that I was eager to return to sunny South Africa, and I took up a postdoc position in the CSIR's robotics group. Of course, this move into robotics led me to the wonderful world of probabilistic reasoning! After joining the Vision and Learning research group at Stellenbosch University I decided to focus on applications of graphical models in computer vision, as well as computer vision problems in autonomous navigation. I've recently begun dabbling in higher-level representation and reasoning, and things like the visual grounding of language, which I think is really cool.
WhAT WILL YOU Be teaching?
Nyalleng Moorosi and I will present an overview of the fundamentals of machine learning. Topics we'll touch on include learning from data, frequentism and Bayesianism, optimisation, generalisation, empirical risk minimisation, and also a few common approaches to supervised and unsupervised learning in both discrete (classification) and continuous (regression) settings.
What advice would you give to those getting started in machine/deep learning?
For me it is truly amazing and inspiring to witness the enthusiasm that so many young South Africans currently have for this field, and also the many great opportunities emerging in local industry and academia. My advice to those entering machine learning would simply be to go for it! I also see an extraordinary collaborative spirit in the ML research community, emphasised by the many open software packages available, which I think has catalysed revolutionary new ideas in a very dramatic way. I encourage you to adopt and promote this amazing spirit!