Catastophic Forgetting in Neural Networks

18 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

iCaRL: Incremental Classifier and Representation Learning (2016, background): “’s survey [27] for classical and Goodfellow et al.”
Neuron clustering for mitigating catastrophic forgetting in feedforward neural networks (2014, background): “Catastrophic forgetting is a well-studied problem that affects artificial neural networks as well as many other supervised learning systems [1].”
Mitigation of catastrophic forgetting in recurrent neural networks using a Fixed Expansion Layer (2013, background): “Mainstream exploration of the problem of catastrophic interference has commonly been restricted to the domain of auto-associative pattern learning [16]; this task compares the ability of a network to retain previously captured sequences of observations after demonstrating the learning of new sequenc”
Controlling Recurrent Neural Networks by Conceptors (2014, background): “Although a number of proposals have been made which partially alleviate the problem in special circumstances ([32, 42, 73], skeptical overview: [77]), catastrophic forgetting was still considered a main challenge for neural learning theory in an expert’s hearing solicited by the NSF in 2007 ([21]).”
Neuron Clustering for Mitigating Catastrophic Forgetting in Supervised and Reinforcement Learning (2015, background): “Catastrophic forgetting is a problem that affects artificial neural networks as well as other learning systems [1].”