About myself

I'm a 6th-year computer science Ph.D. student at the University of Washington, advised by Dan Weld and Mausam. I have been working primarily on designing scalable domain-independent algorithms for planning under uncertainty but have dabbled in other areas as well, including information retrieval, robotics, and first-order probabilistic languages.

Before the Ph.D. adventure, I worked for 2 years at Microsoft's Desktop Search group, and yet before that was busy getting a double B.A. in computer science and applied mathematics at UC Berkeley. While there, I also participated in research on BLOG with Brian Milch and Stuart Russell. Going back to prehistoric times, I finished secondary school #1234 (this is the school's real number) in Moscow, Russia.

Tutorials

• Probabilistic Planning with Markov Decision Processes, slightly different versions of which were given at ICAPS'12 and AAAI'12

Publications

• A Theory of Goal-Oriented MDPs with Dead Ends.
  Andrey Kolobov, Mausam, and Daniel Weld. UAI'12.
  PDF
• LRTDP vs. UCT for Online Probabilistic Planning.
  Andrey Kolobov, Mausam, and Daniel Weld. AAAI'12.
  Accepted for both oral and poster presentation
  PDF
• Reverse Iterative Deepening for Finite-Horizon MDPs with Large Branching Factors.
  Andrey Kolobov, Peng Dai, Mausam, and Daniel Weld. ICAPS'12.
  PDF
• Discovering Hidden Structure in Factored MDPs.
  Andrey Kolobov, Mausam, and Daniel Weld. Artificial Intelligence Journal, May 2012.
  PDF
• Towards Scalable MDP Algorithms.
  Andrey Kolobov, Mausam, and Daniel Weld. IJCAI'11.
  PDF
• Heuristic Search for Generalized Stochastic Shortest Path MDPs.
  Andrey Kolobov, Mausam, Daniel Weld, and Hector Geffner. ICAPS'11.
  PDF
• SixthSense: Fast and Reliable Recognition of Dead Ends in MDPs.
  Andrey Kolobov, Mausam, and Daniel Weld. AAAI'10.
  Accepted for both oral and poster presentation
  PDF
• Classical Planning in MDP Heuristics: With a Little Help from Generalization.
  Andrey Kolobov, Mausam, and Daniel Weld. ICAPS'10.
  Best Paper Award nomination
  PDF
  
  • Early version:
    Determinize, Solve, and Generalize: Classical Planning for MDP Heuristics.
    Andrey Kolobov, Mausam, and Daniel Weld. Workshop on Heuristics for Domain Independent Planning at ICAPS'09
    PDF
• Integrating Paradigms for Approximate Probabilistic Planning.
  Andrey Kolobov. Doctoral Consortium of ICAPS'09.
  Best Paper Award
  PDF
• ReTrASE: Intergating Paradigms for Approximate Probabilistic Planning.
  Andrey Kolobov, Mausam, and Daniel Weld. IJCAI'09
  PDF
  • Early version:
    Regressing Deterministic Plans for MDP Function Approximation.
    Andrey Kolobov, Mausam, and Daniel Weld. Workshop on A Reality Check for Planning and Scheduling Under Uncertainty at ICAPS'08
    PDF
• BLOG: Probabilistic Models with Unknown Objects.
  Brian Milch, Bhaskara Marthi, Stuart Russell, David Sontag, Daniel L. Ong, and Andrey Kolobov. IJCAI'05
  PDF
  • Book Chapter:
    BLOG: Probabilistic Models with Unknown Objects.
    Brian Milch, Bhaskara Marthi, Stuart Russell, David Sontag, Daniel L. Ong, and Andrey Kolobov. Introduction To Statistical Relational Learning (edited by Lise Getoor and Ben Taskar), The MIT Press, 2007.
    PDF
  • Technical Report:
    BLOG: First-order Probabilistic Models with Unknown Objects
    Brian Milch, Bhaskara Marthi, Stuart Russell, David Sontag, Daniel L. Ong, and Andrey Kolobov. UC Berkeley, 2004.
    
• Approximate Inference for Infinite Contingent Bayesian Networks.
  Brian Milch, Bhaskara Marthi, David Sontag, Stuart Russell, Daniel L. Ong, and Andrey Kolobov. AISTATS'05
  PDF
• Improving BLOG Inference with Rao-Blackwellization.
  Andrey Kolobov. Computer Science Honours Program Report, UC Berkeley, 2004.
• Pruning Fuzzy Ontologies.
  Andrey Kolobov, Daniel Kuo, Martine De Cock, and Masoud Nikravesh FLINT-CIBI International Joint Workshop on Soft Computing for Internet and Bioinformatics 2003.

Software

• Glutton, a solver for large finite-horizon MDPs with high branching factors.
• Gourmand, a further development of Glutton.