Learning to Learn Neural Networks
Machine Learning
2016-10-20 v1 Machine Learning
Abstract
Meta-learning consists in learning learning algorithms. We use a Long Short Term Memory (LSTM) based network to learn to compute on-line updates of the parameters of another neural network. These parameters are stored in the cell state of the LSTM. Our framework allows to compare learned algorithms to hand-made algorithms within the traditional train and test methodology. In an experiment, we learn a learning algorithm for a one-hidden layer Multi-Layer Perceptron (MLP) on non-linearly separable datasets. The learned algorithm is able to update parameters of both layers and generalise well on similar datasets.
Cite
@article{arxiv.1610.06072,
title = {Learning to Learn Neural Networks},
author = {Tom Bosc},
journal= {arXiv preprint arXiv:1610.06072},
year = {2016}
}
Comments
presented at "Reasoning, Attention, Memory" workshop, NIPS 2015