Computer Science Seminar (Online)
Date: Tuesday, 22nd March
Time: 2:00 PM to 3:30 PM (online)
Bandit Optimization beyond linear losses + Research opportunities at Google Research India
Research Scientist, Google India.
This talk will be roughly of 40 min duration -- 25 mins for a technical talk and 15 mins for a broad overview of research opportunities for students at Google Research India. We will then have a 20 mins Q&A session with final year or pre-final year undergraduates as well as Masters' students. Finally, we will have a 30 mins Q&A session exclusively for female students and other historically underrepresented groups, including and not limited to, women with disabilities and those from the LGBTQIA+ community. We invite participation from final year or pre-final year undergraduates as well as Masters' students across these demographic groups for this session.
Bandit optimization is a sequential decision making problem where a learner repeatedly interacts with an environment and receives a loss for its action. The goal of the learner is to choose actions which minimize the cumulative loss it suffers. Bandit optimization has received a lot of attention in recent years owing to its applications in a wide range of problems such as design of clinical trials, market pricing, black-box optimization, and recommender systems. Despite its popularity, efficient algorithms for bandit optimization are only known for linear loss functions. In this work, we move beyond linear losses and consider quadratic loss functions. Our main contribution is to design a computationally efficient algorithm that achieves optimal regret in this setting. Our algorithm is a bandit Newton style algorithm which achieves the right balance between exploration and exploitation.
Overview of research opportunities for students at Google Research India
Google research India has several opportunities for graduating/final year students to participate in research projects. Here is a partial list:
1. Predoc researcher: Candidates who hold Bachelor's/Master's degrees can spend up to two years working with research teams at Google research India on cutting-edge research projects. Most of our past pre-doc researchers have then gone on to pursue PhDs at top schools such as MIT, Berkeley, CMU, University of Washington etc.
2. Student researcher/ student internships: This is for students who are in their final/pre-final year of their Bachelor's/Master's or any year of their PhD program to spend part of their time working on research projects with teams at Google research India.
Arun Sai Suggala is a Research Scientist at Google research India. He is broadly interested in online learning, game theory, and their applications to statistical problems such as robustness, boosting. His work has received the best student paper award at ALT'20. Arun obtained his PhD in Machine Learning from Carnegie Mellon University (CMU), working with Pradeep Ravikumar. Prior to CMU, he completed his undergraduate studies in Computer Science and Engineering in the Indian Institute of Technology, Bombay.