John Healy is a senior researcher at the Tutte Institute for Mathematics and Computing in Ottawa, Canada. His recent research interests are in unsupervised learning, ranging from clustering, to dimension reduction, to text analysis. His focus in most of his research is in developing useful techniques with strong mathematical underpinnings that are robust to both noisy data and human error.
In conjunction with Leland McInnes he has developed the popular HDBSCAN library included in scikit-learn-contrib as well as the more recent dimensionality reduction technique called UMAP.
Thursday 1:30 PM–2:15 PM in Track 3 Room