From f76ac4e00ef06e8e6811566c1212a13e57af56ae Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?David=20Mart=C3=ADnez?= <dmartinez@iri.upc.edu> Date: Tue, 12 Jul 2016 14:19:29 +0000 Subject: [PATCH] ReadMe.txt edited online with Bitbucket --- ReadMe.txt | 29 +++++++++++++++++++++-------- 1 file changed, 21 insertions(+), 8 deletions(-) diff --git a/ReadMe.txt b/ReadMe.txt index 0698d65d..6c37401a 100644 --- a/ReadMe.txt +++ b/ReadMe.txt @@ -25,6 +25,7 @@ If you have any question or suggestion, don't hesitate to contact me: email: dmartinez@iri.upc.edu + =========================== === Documentation === =========================== @@ -32,15 +33,21 @@ Online documentation can be found at: www.iri.upc.edu/people/dmartinez/rexd_docs/index.html + ====================== === Introduction === ====================== -REX-D is a relational reinforcement learning with demonstrations algorithm. +REX-D is a relational reinforcement and active learning algorithm. -When using this code, please cite: +When using this code, if using REX-D please cite: Relational reinforcement learning with guided demonstrations, David MartÃnez, Guillem Alenyà , Carme Torras. Artificial Intelligence, 2016 +If using the domain model learner, please cite: +Learning Relational Dynamics of Stochastic Domains for Planning, David MartÃnez, +Guillem Alenyà , Carme Torras, Tony Ribeiro and Katsumi Inoue. In proc. of the +International Conference on Automated Planning and Scheduling, 2016, pp. 235-243 + ====================== @@ -121,9 +128,12 @@ Copy the learner to the bin folder $ cp bin/LFIT ../../bin/lfit -====================== -=== REX-D Usage === -====================== + +============================== +=== REX-D / V-MIN Usage === +============================== +Use RL with a teacher to solve a problem (and learn it). + Go to the scenarios folder. $ cd scenarios @@ -140,9 +150,11 @@ $ ../bin/simulator.sh -====================== -=== Learner Usage === -====================== +=================================== +=== Domain Model Learner Usage === +=================================== +Learn a domain model from a set of input experiences. + Go to the scenarios folder. $ cd scenarios @@ -165,6 +177,7 @@ $ bin/learner_utils evaluate_rules learned_rules.ppddl 2000 Note that the provided version of LFIT can be very slow with the Elevators domain. Our aim is to have a faster one in the short future. + ====================== === New domains === ====================== -- GitLab