Hey there, I am Chandra!

Fifth-year CS PhD student at the University of Southern California, working with Prof. Jiapeng Zhang.

My primary research interest is in design and analysis of algorithms for high-dimensional and large-scale data, currently focusing on genomics and image datasets.

In my first two years if PhD, I focused particularly in unsupervised learning, approaching it from a random matrix theoretic perspective, focused on community detection in stochastic block model. See research for more details.

In the last two years, I have focused on characterizing and exploiting structures in real-world data. Specifically, we have observed some very interesting density-geometry correlations in large scale genomics and image datasets that have shown promise in design of efficient unsupervised, self-supervised, and semi-supervised learning algorithms. See On density-geometry correlations for more details.

Updates

We have released our new package cplearn, a novel unsupervised inference tool for layer-extraction, clustering, and visualization. More updates to follow.

Teaching

Besides my research, teaching is my favorite aspect of pursuing a PhD. I have had the opportunity to be a TA for various courses during my time at USC, as well as take part in K-12 outreach and REU (more details in teaching). I look forward to teaching a full course at some point in the near future.