PhD student in causal modelling at University of Edinburgh

Hi, I’m Steph. You’ve probably heard that correlation does not equal causation. But causality is central to the way we think. Causal modelling is a rigorous way to formalise causality and return notions of cause and effect to science through the tools of machine learning.

My main research is broadly in computational cognitive science and more narrowly in causal modelling, unifying the fields of causal inference and causal selection. These are similar tasks that use directed acyclic graphs and probability theory, including Bayes, as key abstract representations of relationships between variables in the world. The representations can then be interrogated for answers.

I built a model that unifies these two fields (see first paper in the Publications section). Now I’m trying to apply these skills to solving complex problems for businesses. Please email (sidebar) for a meeting to discuss.

Causal matters and DAGs pop up everywhere as a useful method to get your data and analysis pipeline straight even in areas that aren’t explicitly about causes, including choosing the right predictors in regressions. It all depends on getting clear about what is the data generating process. I can help you do that.

I’m funded by the centre for doctoral training for Natural Language Processing. That means I have picked up a huge stash of passive knowledge about large language models and issues of syntax and semantics, even though these are not my main area.

I am also pretty much obsessed with mountains (hill running, trail ultras, climbing) and Buddhist meditation.