Bayesian Structure Adaptation for Continual Learning

By Abhishek Kumar, Sunabha Chatterjee, Piyush Rai. 2019

35 citations recorded by the source corpus.

Continual Learning is a learning paradigm where learning systems are trained with sequential or streaming tasks. Two notable directions among the recent advances in continual learning with neural networks are ($i$) variational Bayes based regularization by learning priors from previous tasks, and, ($ii$) learning the structure of deep networks to adapt to new tasks. So far, these two approaches have been orthogonal. We present a novel Bayesian approach to continual learning based on learning the structure of deep neural networks, addressing the shortcomings of both these approaches. The proposed model learns the deep structure for each task by learning which weights to be used, and supports inter-task transfer through the overlapping of different sparse subsets of weights learned by different tasks. Experimental results on supervised and unsupervised benchmarks shows that our model performs comparably or better than recent advances in continual learning setting.

Subfields: Artificial Intelligence, Meta & Continual Learning

How researchers see this paper

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

In Artificial Intelligence

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “The dynamic architecture can be explicitly optimized to improve parameter efﬁciency and knowledge transfer, such as by reinforcement learning [229], [230], architecture search [230], [231], variational Bayes [232], etc.”
Recent Advances of Continual Learning in Computer Vision: An Overview (2021, neutral): “[144] proposed to build each hidden layer with the Indian Buffet Process [145] prior.”
How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition (2022, methodology): “In particular, Zhang et al. (2021) developed a continual variational inference method using a Chinese restaurant process to automatically determine the number of latent components, while Kumar et al. (2021) did the same using an Indian buffet process for unsupervised and supervised learning.”
VariGrow: Variational Architecture Growing for Task-Agnostic Continual Learning based on Bayesian Novelty (2022, methodology): “For a Bayesian treatment of ω , approximate Bayesian inference techniques for neural networks, such as Monte Carlo dropout and its variants can be used (Gal & Ghahramani, 2016; Gal et al., 2017; Kingma et al., 2015; Boluki et al., 2020; Kumar et al., 2021).”