I am interested in decision-making and learning under uncertainty. I focus on methodologies and techniques that combines human judgement, published evidence and uncertain data to improve learning. This includes the use of Bayesian approaches, causal and probabilistic graphical models for learning quantitative ‘white-box’ models that can be understood and revised by domain experts. I am also interested in knowledge-representation and abstraction techniques in probabilistic models to develop decision support tools that reasons with different level of abstractions in a similar way to human experts. I apply my research to decision-making problems in different disciplines including health-care and agriculture.