I am currently working as a Visiting Research Scholar under Dr. Judea Pearl researching Autonomous Agents and Learning Systems.
Prior to that, I worked with various startups across U.S., U.K. and India in business intelligence roles for dev teams to better understand user behavior, growth and engagement. My academic background is in Mathematics with minors in Statistics and Computer Science.
My primary interests are in MultiAgent Systems, AutoML, Procedural Content Generation, and Reinforcement Learning. I am concurrently also building interests in Cognitive Sciences and Hierarchical Learning.
I am, foremost, driven by the goal to design personalized large-scale systems to replicate three key behavioral aspects of human decision making - Activities, Goals and Interactions- AGI. They key question that motivates me is why we do that we do and if/could you reverse engineer and manipulate it like an input/output pair.
I have been teaching machine learning for around 2.5 years for several organizations including Women Who Code, WiMLDS in an effort to give back to the community.
- July 30: Causality "is" Correlation: Back to Fundamentals paper available
- July 20: Commentary Paper on using Causal Inference in Reinforcement Learning available
- July 15: Novelty Search in Machine Learning Commentary Paper Available
- July 10: Problems & Challenges in AutoML Pre-Print Available, joint work with Shubhi Sareen
- June 27: Cooperation v/s Competition in MultiAgent Systems Paper Submitted to AIIDE
- June 22: Blogpost on Games for A.I. Available. (Working Draft)
- June 2019: Attending ICML 2019 in Long Beach
- October 2018: Committee Co-Chair for PyData Los Angeles where I presented a talk on Big Problems at the Heart of Machine Learning and a tutorial on
A Hands-On Application of Causal Methods in Python
- June 2018: Competed in AutoML for LifeLong Machine Learning Challenge for NeurIPS 2018
- I am competing in Animal-AI Olympics for NeurIPS 2019.
- On December 5, 2018 I started MLNerdie to solve the induction problem in Machine Learning. I conduct weekly Saturday classes in Deep Reinforcement Learning
If you happen to be interested in the same topics and are looking to collaborate, please feel to email me with a rough 1-page proposal that includes the abstract and some background readings you have conducted on the topic.
Also, I am always happy to visit labs or companies to give talks - shoot me an email if you'd like to discuss further!