Research

My graduate research in computational social science focused on a multidisciplinary approach, leveraging advanced statistical and computational techniques to delve into the complexities of behavior diffusion and its implications in real-world applications. I utilized a broad array of analytical tools, including point processes, Bayesian modeling, convex optimization, causal inference, graph theory, survival analysis, and machine learning. My primary area of study revolved around the utilization of information diffusion models for predicting real-world events, specifically in the realms of cascading network processes and cybersecurity, and an examination of the circumstances under which such models may fail to provide accurate predictions. Additionally, I dedicated a significant part of my research to designing and analyzing randomized controlled experiments aimed at understanding human preferences in the presence of social influence.

Since then, I have gained extensive experience in the industry, particularly in the realm of information retrieval and recommendation systems. I have worked on machine learning for search ranking systems, designed to handle massive-scale queries in real-time. These systems were developed based on machine learning models trained on terabyte-scale data, while also being customized for seamless deployment to production systems. Furthermore, I've utilized reinforcement learning systems to enhance machine learning models for search results, specifically aiming to eliminate defects from search result pages.

I have recently started focusing on using causal inference for mitigating bias in microlending platforms, and designing bandit based algorithms for matching markets in peer lending. My recent work on reinforcement learning focuses on designing algortihms for handling adversaries in decentralized networked multi-agent environments.

Some of my current interests are:

1. Large scale machine learning (deep learing models, continual learning, information retrieval systems).
2. Behavior diffusion in online social networks.
3. Sequential Decision Making with applications in markets for hiring and lending.
4. Scientific NLP.

Other

I spend quite a large part of my time outside of work on books, movies and swimming. I also occasionally enjoy playing outdoor soccer.