•The Data Science Labs for Differential and Integral Calculus (MA16290) are now published on Github.

•My grad student Evzenie Coupkova just posted our preprint proving that classification by thresholding after projection on a random line is asymptotically optimal and has minimal complexity (i.e. minimal generalization error).

•Our work on shape from echoes is features in the news! Here is the press release from TU Munich, with accompanying video, the press release from Purdue, and a SIAM nugget article (more...)

•“A Drone Can Hear the Shape of a Room,” my new paper with Gregor Kemper, is now up on the SIAM website. See the project page for more information.

•Our manuscript on “A highly likely clusterable dataset with no cluster,”  is now posted on the arXiv. This paper highlights 1) how the choice of projection for reducing dimension can create different clusterings, and 2) a mathematical reason why datasets in a high-dimensional space are likely to cluster by random projection.

• “A Drone Can Hear the Shape of a Room”, my new paper with Gregor Kemper, is now posted on the arXiv. This is another problem involving unlabeled distances.

•Our paper titled “A 3-Step method to estimate phenylalanine in commercial food for PKU management” is now available as a preprint in IEEE Xplore.

• Our paper on “Clusterability and Clustering  of real high-dimensional data” has appeared in IEEE Transactions on Image Processing. 

•Several people wanted to know about the tricks Shanshan Huang and I developed to quickly identify simple shapes in images. So here is a short tutorial that covers three basic tricks:

        Tutorial: “How to quantify shape elongation and how to identify symmetric shapes in a gray scale image”

• Here is another tutorial that explains the relationship between the Pascal Triangle of an image and its Radon transform.

        Tutorial: “Relationship between the Pascal Triangle and the Radon transform of an image.”