2025-05-25 Weekly Notes
For most of last week I attended the Health Impact Assessment workshop organised by the Public Health Modelling Group from the MRC Epidemiology Unit in Cambridge. It was a great in-depth introduction to epidemiological modelling using mostly longitudinal data for health assessment. It is very relevant as the group focuses on the effects of urban features in different health outcomes.
Also, I attended the AI4ER workshop in West Hub where some of my PhD peers presented their work. Worth highlighting is Orlando Timmerman’s work in long-term coral suitability, Joshua Dimasaka’s work in catastrophe modelling using graph neural networks. And shout-out to Sanjoo Paddea and Gilly Walker for organising an amazing day of networking spaces and lovely talks for all AI4ER and adjacent people.
For my project I mostly worked on two things:
- I started labelling nature features on Charles Booth’s maps using Label Studio, focusing on individual trees (broadleaf and conifer) and clusters of trees, as well as water bodies. I met with Hendrik Herold to discuss some ideas on what the best approach is to extract all other features, including poverty and building footprints.
- I worked on my presentation with the SDG group where I will be presenting what I’ve done in the last couple of weeks and updates on my paper, for which I designed the plot explaining what I had drawn in a previous post about Gini coefficient and measuring tree visibility (see below).
Visualization
