Supplementary Material for "The Effects on Adaptive Behaviour of Negatively Valenced Signals in Reinforcement Learning"

This item is shared privately

dataset

modified on 2018-08-22, 13:28

zip files with raw data from the 10 random initializations of the 4 best hyperparameter sets per condition and per activation function. Organized by function and per condition.

pdf file containing the pair-wise comparison of the three tested conditions against the baseline.

Link to the supplementary material of the original paper, includes matlab command and data set for learning:

https://figshare.com/s/becd4366e4ed38731353

References:

Navarro-Guerrero, N., Lowe, R., & Wermter, S. (2017). The Effects on Adaptive Behaviour of Negatively Valenced Signals in Reinforcement Learning. In Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EPIROB) (pp. 148–155). Lisbon, Portugal: IEEE. https://doi.org/10.1109/DEVLRN.2017.8329800

Navarro-Guerrero, N., Lowe, R., & Wermter, S. (2017). Improving Robot Motor Learning with Negatively Valenced Reinforcement Signals. Frontiers in Neurorobotics, 11(10). https://doi.org/10.3389/fnbot.2017.00010

Contact:

For information please see the paper. For specific questions regarding the paper please contact Nicolás Navarro-Guerrero.

Keywords

reinforcement learning Inverse kinematics (IK)punishment and nociception Knowledge Representation and Machine Learning

Licence

CC BY 4.0

Supplementary Material for "The Effects on Adaptive Behaviour of Negatively Valenced Signals in Reinforcement Learning"

Categories

Keywords

Licence