Michael Katz

Michael Katz

Home Publications Co-authors Activities CV (pdf) Software Contact Google Scholar IBM Research web site GitHub page

Software

The page includes links to the software I was partially responsible for developing.

K* search based planners for top-k/top-quality planning also pip-installable via pip install kstar-planner

Forbid-Iterative (FI) Planner suite for top-k, top-quality, and diverse computational tasks also pip-installable via pip install forbiditerative

Diversity score computation for a set of plans

Cerberus planner, post-IPC 2018 version

IBM Research AI Planning Service

Competitions

I have participated in the deterministic part of the IPC-2018 - International Planning Competition 2018
with multiple planners. One of them, Delfi has won the sequential optimal track. The source code for Delfi can be found here.

Other planners from IPC2018 include

Metis, IPC 2018 version

MERWIN, IPC 2018 version

Cerberus, IPC 2018 version

Mercury 2014, IPC 2018 version

I have participated in the deterministic part of the IPC-2014 - International Planning Competition 2014
with two planners, Mercury for the sequential satisficing track and Metis for the sequential optimal track.
Mercury (with Joerg Hoffmann) has won two awards:

Runner-Up in the sequential satisficing track, and

Innovative Planner Award.

The source code for the planners can be found here (or by request):

Mercury

Metis

Older

I have participated in the deterministic part of the IPC-2011 - International Planning Competition 2011
with four planners based on Implicit Abstraction Heuristics.

Sequential Satisficing Track

ForkUniform - Planner iteratively runs Weighted A* with Fork Decomposition based heuristic.

Sequential Optimal Track

ForkInit - Planner runs A* with Fork Decomposition based heuristic, optimal for initial state action-cost partitioning.
IForkInit - Planner runs A* with Inverted Fork Decomposition based heuristic, optimal for initial state action-cost partitioning.
LMFork - Planner runs LMA* with heuristic calculated on the landmarks enriched task, Fork Decomposition based heuristic, optimal for initial state action-cost partitioning.

The code of Implicit Abstraction Heuristics is implemented on the Fast-Downward platform, FDTech branch. The code is available upon request.