IBaCoP (IPC-2018)

We trained a predictive model for a (yes/no) classification task using Rotation Forrest (Rodriguez, Kuncheva, and Alonso 2006). The model tries to encode whether a given planner will solve the planning task or not. IBaCoP2-2018 is the result of querying this model and selecting the five planners with the best “positive” prediction confidence.

Planners:

• jasper • mercury • BFS(F) • SIW • FDSS-2 • probe • yashp2-mt • lama-2011 • lamar • arvand

Planner abstract: https://icenamor.github.io/files/IBaCoP_2018.pdf

Source-code: git clone https://bitbucket.org/isabelcenamorg/ibacop2-2018-final-version.git