Statistical Modelling with Graphical Models
Sofia Massa (University of Oxford)
Friday 11th October, 2013 15:00-16:00 Maths 204
Graphical models have been studied and formalised across many communities of researchers (artificial intelligence, machine learning, statistics, to name just a few) and nowadays they represent a powerful tool for tackling many diverse applications. They still represent an exciting area of research and many new types of graphical models have been introduced to accommodate more complex situations arising with more challenging research questions and data available. Even the interpretation of graphical models can be quite different in different contexts. If we think for example of high dimensional settings, the original notion of conditional independence between random variables of a graphical model and encoded by the conditional dependence graph is generally lost and the interest is in finding the most important components of thousands of random variables.
In this talk we will present some of the challenges we are faced when working with graphical models in applied contexts. We will go through some case studies arising in different areas of applications (genetics, experimental psychology, meta-analysis, among others) trying to elucidate why graphical models are still an important tool for researchers and what are some of the new frameworks arising. In particular we will show how graphical models can be very powerful for both an explorative statistical analysis and for more advanced statistical modelling and prediction.