Michael Katz  Michael Katz photo       

contact information

Mobile Analytics - Automated Decision Making
Haifa Research Lab, Haifa, Israel

links

Professional Associations

Professional Associations:  Association for the Advancement of Artificial Intelligence (AAAI)

profile


NEWS: My planner Mercury (with Joerg Hoffmann) has won two awards at the 2014 International Planning Competition:

  1. Runner-up of the Deterministic Sequential Track 
  2. Innovative Planner Award

 

Before joining IBM I have spent two years in France and Germany, doing a postdoc hosted by Prof. Joerg Hoffmann at the Institut national de recherche en informatique et en automatique (INRIA), Nancy, France and in the Department of Computer Science, Saarland University, Germany.
Before that I was a postdoc at the Technion - Israel Institute of Technology, in the Faculty of Industrial Engineering & Management, where I also did my PhD.

My PhD studies were done in the field of Artificial Intelligence. My PhD Thesis Implicit Abstraction Heuristics for Cost-Optimal Planning was the winner of the ICAPS Best Dissertation Award 2011.
A brief summary of my PhD Thesis was published in AI Communications Journal. These two pages of comparatively light reading present the general idea of my thesis.


Currently my focus has somewhat shifted to using classical planning to solve real life applications.

In general, the main focus of my work is classical automated planning, both satisficing and cost-optimal. The prefered method of solving classical planning problems is heuristic search in the problem's state space. My main work focused on automatic derivation of heuristic functions for classical planning.

In general, I am interested in several adjacent areas, such as Autonomous systems and Artificial Intelligence (AI), general (domain independent) problem solving, planning and planning complexity, constraint satisfaction and optimization, combinatorics and graph algorithms. Also, I am interested in making planning accessible to software developers by simplifying problem modelling.