Overcoming catastrophic forgetting in neural networks

James Kirkpatrick

[0]

[0]

[0]

[0]

[0]

[0]

Kieran Milan

John Quan

Tiago Ramalho

[0]

Agnieszka Grabska-Barwinska

[0]

Demis Hassabis

[0]

Claudia Clopath

[0]

Proceedings of the National Academy of Sciences of the United States of America, Volume abs/1612.00796, Issue 13, 2017.

Cited by: 1766|Views664

Keywords:

artificial intelligencecontinual learningdeep learningstability plasticitysynaptic consolidation

Weibo:

We propose a dedicated continual learning method for graph neural networks, which is to our best knowledge the first attempt along this line

Abstract:

The ability to learn tasks in a sequential fashion is crucial to the development of artificial intelligence. Until now neural networks have not been capable of this and it has been widely thought that catastrophic forgetting is an inevitable feature of connectionist models. We show that it is possible to overcome this limitation and train...More

Code:

https://github.com/hhliu79/TWP

Data:

Full Text

View via Publisher