Zoubin Ghahramani
University of Cambridge
- Research field
- Probabilistic machine learning, Bayesian methods, Gaussian processes
Zoubin Ghahramani is Professor of Information Engineering at the University of Cambridge and one of the founders of the modern probabilistic view of machine learning, in which everything a system is unsure about is represented as a probability distribution and learning is the computation of a posterior. His 2015 review in Nature sets out that view and four lines of research built on it, among them probabilistic programming and Bayesian optimization. The unit presents the review as the programme of which Bayesian networks, the naive Bayes classifier, and hidden Markov models are the first step.
In the masterclass
Section 2, after inference by enumeration, presents his review as the argument for machines that represent their own uncertainty.
The research is presented in Unit 3: Reasoning under Uncertainty, in a research spotlight box that explains the result at the level of the unit and links it to what you have just learned.