Directly comparing image pixels to decide iftheybelong tothe same cluster leads to poor results because they are strongly impacted by factors irrelevant to clustering, such as exact viewpoint or lighting.

Our study begins with an observation that the BW metric has alinear dependence on SPD matrices in contrast tothe quadratic dependence ofthe AI metric.

We explore the minimax optimal error associated with a demographic parity-constrained regression problem within the context of a linear model.

There is an increasing need for effective active learning algorithms that are compatible with deep neural networks.

The above problem requires the knowledge of the distributionπ which is typically unknown in practice.