I am a final year Electrical Engineering student at IIT Kanpur. I am enrolled in a 5-year integrated Masters’ programme. The primary object of my interest is Statistical Machine Learning, particularly Bayesian ML.

In the fall of 2018, I did a 6-month stint at the Max Planck Insitute for Intelligent Systems(MPI-IS) in Tuebingen, Germany. I was fortunate enough to be supervised during this period by Dr Isabel Valera, and worked alongside Dr Antonio Vergari, who taught me several tricks of the trade in research :). My work in MPI-IS was primarily focused on Bayesian Nonparametrics and Hawkes Processes. You can find our work in the publications page.

During the Summer of ‘19, I interned in the Department of CS at Aalto University in the Computational Systems Biology Group. Here I was supervised by Dr Harri Lahdesmaki and work with Dr Markus Heinonen and Çağatay Yıldız. My work here revolved around designing scalable, higher-order, gradient-based MCMC samplers for inference in ODE systems, which are parameterized by Neural Networks or Gaussian Processes.

Currently for my masters thesis, I am working with Dr Vipul Arora on employing deep genreative models and normalizing flows for controlled and accelerated simulation in the XY model, a model used in statistical physics.

Earlier, I worked on term-projects for courses such as Statistical Learning Theory, Bayesian Machine Learning, Introduction to NLP, Convex Optimization, and Data-Driven Program Analysis. Reports of some of these are available in the Reports section of the publications page.

My interests, though concentrated around Bayesian ML, have been constantly evolving. My current research interest is in developing:

  1. scalable Bayesian inference, particularly gradient-based MCMC techniques,
  2. to get accurate uncertainty estimates for complex data like time series and structured data,
  3. while employing flexible models, such as nonparametric Bayesian models and continuous-time models.

If you’re interested in collaboration with me, feel free to contact me! I plan to start posting occasional blog posts on topics that interest me, so you might want to check up on the site from time to time :).