.

"There is nothing more practical than a good theory" - Kurt Lewin

I was a PhD student in Computer Science at Rutgers. I successfully defended my dissertation in March 2014 and I am now working on optimization algorithms for advertising at Google in New York.

At Rutgers I worked with Prof. Michael Littman and Prof. Haym Hirsh. I am very interested in all the sets from the following inclusions:
Artificial Intelligence ⊃ Machine Learning ⊃ (this inclusion is debatable) Reinforcement Learning ⊃ {Bandit algorithms, Monte Carlo Tree Search methods, Markov Decision Processes}. More details: publications and reports.
I am also inspired by ideas from Computational Learning Theory and Computational Complexity. If I could live a parallel life, I'd love to do research in Physics.
I come from Medgidia, a charming town in beautiful Dobrogea, Romania.

News

March 2014 I started a new job as a Software Engineer at Google NY in the Ads Optimization group.

March 2014 I successfully defended my PhD dissertation, "Stochastic Dilemmas: Foundations and Applications" (to appear). A big thank you to my thesis committee: Michael Littman (adviser), Haym Hirsh (adviser), Swastik Kopparty and Shie Mannor (external member) for all their help - it's been a great journey!

Spring 2013: I was a TA for CS 536 (Machine Learning).

Fall 2012: I passed my PhD qualifier exam. A big thank you to the committee members: Michael Littman, Haym Hirsh, Swastik Kopparty and Muthu Muthukrishnan for their feedback! Here are the slides of my presentation.

Summer 2012: I was a Software Engineering Intern at Google NY in the Ads Optimization group.

October 2011: My work with Ari Weinstein, Michael Littman and Erick Chastain won the best student paper award at the NY Machine Learning Symposium. We showed how to derive upper & lower bounds in a model inspired by the PAC Bandit framework with the goal of doing efficient open loop planning in 'hard' domains - read video games :). The video of the talk.

Summer 2011: I was a Software Engineering Intern at Google NY in the Google Goggles team.

December 2010: I will help organize a workshop on Monte Carlo Tree Search methods at ICAPS'11: http://icaps11.icaps-conference.org/workshops/mcts.html

Fall 2010: I am a Research Assistant in the RL3 Lab.

Summer 2010: I was a Software Engineering Intern at Google NY in the Google Crisis Response team.

Spring 2010: I was a TA for CS 205 - Introduction to Discrete Structures (class, recitation).

Summer 2009: I was a TA for CS 205 - Introduction to Discrete Structures (class, recitation).

June 2009: RU-Poly team (Brian Schubert, Ravneet Singh, Yang Xiong and myself) finished RL-competition 2009, Polyathlon domain on the 3rd place. We finished the "regular season" in the 1st place. The competition was fun, the software was easy to use, the forum admins were very responsive. In a word, a very nice experience.

Fall 2008, Fall 2009: I was a TA for CS 105 (Great Insights in Computer Science). It's a really cool class and I warmly recommend it to all non-CS majors.

Grad classes so far:

  • Computer Science: Advanced Topics in Computational Learning Theory (my scribe notes for the class), Computational Geometry, Probabilistic Algorithms, Computational Complexity, Learning and Sequential Decision Making, Principles of Knowledge Representation and Reasoning, Pattern Recognition, Design and Analysis of Algorithms
  • Statistics: Statistical Learning Theory and High Dimensional Data Mining, Theory of Statistics, Theory of Probability
  • Mathematics: Analysis of Boolean Functions, Combinatorics 1