Speaker

Talk Title
Inferential Machine Learning: Statistics and Machine Learning for real-world problems
Talk Summary
Statistics and Machine Learning have undoubtedly proved to be useful in tackling problems in a variety of application areas including social science, public health, environmental science, media, or transportation. While these two scientific disciplines share a great number of tools, they have some fundamental differences. Nevertheless, the synergies between them provide a promising avenue in delivering truly useful, explainable and efficient methodologies that can be used in impactful problems.
In this talk I aim to show some examples related to my research, of how Statistics and Machine Learning have been used to answer real-life strategic questions. The first concerns the recent pandemic – one of the biggest challenges we had to face as humanity. Using Covid-19 case and death data we developed methods to characterise the evolution of the pandemic in Brazil, influencing public policy measures. However, the quality of the results relies on the quality of the data. This is another example I will focus on as it is another problem that concerned us during the pandemic: how to correct for delays in data reporting. Secondly, I will talk about approaches for modelling data arising in a social science context with particular focus on crime data. This is a synergy example between Statistics and Machine Learning delivering a flexible model that can explain and predict how events cluster in space and time.