nihars [at]
Office: GHC 8211

Google Scholar page

I am an assistant professor at CMU in the Machine Learning and the Computer Science departments. I work in the areas of statistical learning theory and game theory, with a focus on learning from people.

I recently completed my PhD in the EECS department at UC Berkeley, working with Martin Wainwright and Kannan Ramchandran. My thesis committee members were Martin Wainwright, Kannan Ramchandran, Christos Papadimitriou, and Tom Griffiths. In the summers of 2013 and 2014, I interned at Microsoft Research, Redmond working with Denny Zhou in the machine learning group.

Recent news

My research interests lie in the areas of statistics, machine learning, information theory and game theory, with a focus on the application to crowdsourcing and human-centered systems. Here is a brief description of some of my work.
Statistical inference from crowdsourced data

There is a long and rich literature on learning from data from people. However, the models assumed therein make quite restrictive parameter-based assumptions---they assume that the behavior of any entity (a user, item, etc.) is governed by a single parameter. Although there are algorithms that work well when these assumptions are satisfied, such algorithms fail under violations of these limiting assumptions, which occurs frequently in practice. In my research, I propose a departure from this conventional approach, towards a paradigm of far richer permutation-based models. Under this new approach, I design provably optimal and robust algorithms for estimation, prediction and ranking in human-centered applications. Moreover, I prove that as compared to the parameter-based algorithms, my algorithms enjoy a surprising win-win in the statistical bias-variance tradeoff.

Selected papers:

"Unique" mechanisms for obtaining high-quality data from crowdsourcing

My research also targets robust collection of data via game-theoretic incentives. I have established a framework towards a novel principled approach for the design and selection of incentive mechanisms. I employ this framework to construct incentive mechanisms, and prove that my proposed multiplicative mechanisms are the one and only possible incentive mechanisms in various applications.

Selected papers:

My Google Scholar page


Jingyan Wang, Robotics Institute, CMU
Ivan Stelmakh, Machine Learning Department, CMU (advised jointly with Aarti Singh)