Anuran Makur

About Me

I am a postdoctoral researcher in the Laboratory for Information and Decision Systems (LIDS) and the Institute for Data, Systems, and Society (IDSS) at Massachusetts Institute of Technology (MIT), where I am hosted by Prof. Ali Jadbabaie and Prof. Devavrat Shah.

I completed my Sc.D. (doctorate) from the Department of Electrical Engineering and Computer Science (EECS) at MIT, where I was supervised by Prof. Yury Polyanskiy in LIDS and Prof. Lizhong Zheng in the Claude E. Shannon Communication and Network Group within the Research Laboratory of Electronics (RLE). Previously, I completed my undergraduate work in the Department of Electrical Engineering and Computer Sciences (EECS) at the University of California, Berkeley, where I worked with Prof. Venkat Anantharam.

I am broadly interested in theoretical EECS, statistics, and related applied mathematics problems. In particular, my research interests are in:

  • Theory of Machine Learning
    • ranking and preference learning
    • non-parametric density estimation and regression
    • gradient-based optimization methods for empirical risk minimization

  • Combinatorial Statistics
    • reconstruction/broadcasting on graphs
    • probabilistic cellular automata
    • reliable computation using noisy circuits

  • Information Theory and Statistical Inference
    • information (and functional) inequalities and information contraction
    • modal decompositions and feature extraction for high-dimensional inference
    • fundamental limits of permutation channels

Recently, I have been studying minimax estimation of rankings and skill distributions from pairwise comparison data, inexact gradient methods for optimization in machine learning settings, reconstruction problems on regular grids and closely related questions concerning the ergodicity of associated probabilistic cellular automata, and the information capacity of noisy permutation channels.