Hi, welcome to my webpage!
I am an algorithmist, and work in the intersection of theory and practice. I am interested in developing "extremely simple", practical approximation algorithms with provable performance guarantees for real-world problems. Among other things, my current research focuses around sketching/dimensionality reduction algorithms. In particular, I'm focusing on: a) developing new sketching/dimensionality reduction algorithms for various data types and similarity measures, b) improving existing sketching/dimensionality reduction algorithms by making them fast, scalable and accurate, and c) exploring applicability of such results in various machine learning tasks such as learning node embedding in large scale network, itemset mining, model compression etc.
I generally use techniques from matrix algebra, sampling, random projection, and randomized hashing.
I earned my Ph.D. in Theoretical Computer Science from Chennai Mathematical Institute (Aug 2009 - Nov 2014).
Previously, I have been working with industry research labs in the areas of algorithms in Data Science and Machine Learning at Wipro-AI Research, Bangalore; and TCS Innovation Labs, New Delhi. I have also spent a few months at IIIT Bangalore as a Research Associate.
Openings: I'm looking for highly motivated MS/PhD students to work on the areas mentioned above.
Note: I am not taking students for any short term internships (summer/winter). Unfortunately, I will be unable to individually respond to such emails/queries. Please contact me only if you are interested in at least a year long project, and have strong academic credentials.
My Erdös Number is 3.
Our paper “Variance Reduction in Frequency Estimators via Control Variates Method” joint work with Raghav Kulkarni, has been accepted in the 37th Conference on Uncertainty in Artificial Intelligence (UAI, 2021) -- a core A* ranked conference in Computer Science.
Our paper "Improving Tug-of-War sketch using Control-Variates method" joint work with Bhisham Dev Verma and Raghav Kulkarni has been accepted in SIAM Conference on Applied and Computational Discrete Algorithms (ACDA21).
Our following two papers got accepted in ACML 2020
1) Randomness Efficient Feature Hashing for Sparse Binary Data. Joint work with Karthik Revanuru, Anirudh Ravi, Raghav Kulkarni.
2) Scaling up Simhash. Joint work with Anup Deshmukh, Pratheeksha Nair, Anirudh Ravi. This paper was also invited for Special Issue in Springer Nature Computer Science.
Our paper "Subspace approximation with outliers", joint work with Amit Deshpande, has been accepted in The 26th International Computing and Combinatorics Conference - COCOON 2020 -- a core A ranked conference in Computer Science. We won the Best Paper Award at COCOON 2020 for this submission.
Our paper "Robust k-means++", joint work with Amit Deshpande, Praneeth Kacham, has been accepted in Association for Uncertainty in Artificial Intelligence - UAI 2020 -- a core A* ranked conference in Computer Science.
I am serving as a Programme Committee member of "The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases" (ECML-PKDD) 2020.
We (Mukesh Prasad, Rameshwar Pratap, Rajiv Ratn Shah, Weiping Ding, Javier Andreu-Perez, Guandong Xu) are organising a special session on “Feature Extraction and Learning on Image and Text Data” to be held in conjunction with the conference “IEEE International Conference on Systems, Man and Cybernetics (SMC)-2020”. You may consider submitting your paper. Further details are available here.
I am serving as a Programme Committee member of "The Sixth IEEE International Conference on Multimedia Big Data" (IEEE-BigMM) 2020.
Amit Sangroya (TCS Research) and I are co-chairing IEEE BigMM’20 Grand Challenge which is to be held in conjunction with the IEEE International Conference on Multimedia Big Data from September 24-26, 2020 at New Delhi. You may consider submitting your Grand challenge proposal to us. Further details are available here.