In this talk, we will dive into black-box systems, and discuss which interesting machine learning challenges lay ahead.
Over the past 10 years, I have worked on developing deep learning models for wearable EEG, specifically focusing on the toy case of high precision sleep monitoring. The past few years, this has culminated in exciting projects, studying how sleep changes over time for people with psychiatric illness, chronic pain, or NASA and ESA astronauts visiting the International Space Station. During these studies, we have collected a lot of experience with how to work with monitoring black-box systems in the wild.
In this talk, I will share some of these experiences, and discuss which interesting machine learning challenges lay ahead.
Kaare Mikkelsen has a PhD in statistical physics from Aarhus University from 2014. After this, Kaare switched to a career in biomedical engineering, developing analysis pipelines for wearable EEG. Today he is an Associate Professor at the Department of Electrical and Computer Engineering, at Aarhus University, where he heads the Biomedical Machine Learning group, specializing in deploying and developing data science methods for biomedical data, in particular wearable EEG.
