LUKE (Learning Underlying Knowledge of Experts)

LUKE is a software that uses apprenticeship learning technique to learn model parameters of partially observable environments. We can estimate parameters of unknown enviroment by a given demonstration of the task by an expert.

Required software:

Linux 3.2 or later
g++ 4.8 or later
CPLEX (12.4 or later)
Boost C++ library, 1.53 or later
Intel Threading Building Blocks library
NLopt 2.3 or later
Python

Distributed under GPLv3 license.

For details (installation etc.), please refer to README.

luke-0.91.tar.bz2 (2014/5/16)

luke-0.9.tar.bz2 (2014/3/31)

The development of this software is supported by Aihara Innovative Mathematical Modelling Project, JSPS through the "Funding Program for World-Leading Innovative R&D on Science and Technology (FIRST Program)", and JSPS Grant-in-aid for Young Scientists (B) (25730128).