Automated Curriculum Learning for Neural Networks
Alex Graves, Marc G. Bellemare, Jacob Menick, Rémi Munos, Koray Kavukcuoglu
Abstract
We introduce a method for automatically selecting the path, or syllabus, that a neural network follows through a curriculum so as to maximise learning efficiency. A measure of the amount that the network learns from each data sample is provided as a reward signal to a nonstationary multi-armed bandit algorithm, which then determines a stochastic syllabus. We consider a range of signals derived from two distinct indicators of learning progress: rate of increase in prediction accuracy, and rate of increase in network complexity. Experimental results for LSTM networks on three curricula demonstrate that our approach can significantly accelerate learning, in some cases halving the time required to attain a satisfactory performance level.
BibTeX
@InProceedings{pmlr-v70-graves17a,
title = {Automated Curriculum Learning for Neural Networks},
author = {Alex Graves and Marc G. Bellemare and Jacob Menick and R{\'e}mi Munos and Koray Kavukcuoglu},
booktitle = {Proceedings of the 34th International Conference on Machine Learning},
pages = {1311--1320},
year = {2017},
editor = {Precup, Doina and Teh, Yee Whye},
volume = {70},
series = {Proceedings of Machine Learning Research},
month = {06--11 Aug},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v70/graves17a/graves17a.pdf},
url = {https://proceedings.mlr.press/v70/graves17a.html},
abstract = {We introduce a method for automatically selecting the path, or syllabus, that a neural network follows through a curriculum so as to maximise learning efficiency. A measure of the amount that the network learns from each data sample is provided as a reward signal to a nonstationary multi-armed bandit algorithm, which then determines a stochastic syllabus. We consider a range of signals derived from two distinct indicators of learning progress: rate of increase in prediction accuracy, and rate of increase in network complexity. Experimental results for LSTM networks on three curricula demonstrate that our approach can significantly accelerate learning, in some cases halving the time required to attain a satisfactory performance level.}
}