Continual Learning of Context-dependent Processing in Neural Networks

About

Deep neural networks (DNNs) are powerful tools in learning sophisticated but fixed mapping rules between inputs and outputs, thereby limiting their application in more complex and dynamic situations in which the mapping rules are not kept the same but changing according to different contexts. To lift such limits, we developed a novel approach involving a learning algorithm, called orthogonal weights modification (OWM), with the addition of a context-dependent processing (CDP) module. We demonstrated that with OWM to overcome the problem of catastrophic forgetting, and the CDP module to learn how to reuse a feature representation and a classifier for different contexts, a single network can acquire numerous context-dependent mapping rules in an online and continual manner, with as few as $\sim$10 samples to learn each. This should enable highly compact systems to gradually learn myriad regularities of the real world and eventually behave appropriately within it.

Guanxiong Zeng, Yang Chen, Bo Cui, Shan Yu · 2018

Related benchmarks

Task	Dataset	Result
Continual Learning	Split CIFAR-100 20 tasks	Mean Test Accuracy68.47	62
Image Classification	CIFAR-100 Split	Accuracy77.07	61
Class-incremental learning	CIFAR10 (test)	Average Accuracy75.39	59
Continual Learning	CIFAR-100 (10-split)	ACC68.89	54
Class-incremental learning	MNIST (test)	Average Accuracy96.3	35
Continual Image Classification	CIFAR100 Split	Accuracy50.94	35
Continual Learning	TinyImageNet 25-split	ACC49.98	29
Image Classification	MNIST Split	Test Accuracy99.36	24
Class-incremental learning	Tiny-ImageNet standard (test)	Average Accuracy40.29	20
Continual Learning	MNIST permuted	AT90.71	19