Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors

Nikhil Mehta, Kevin Liang, Vinay Kumar Verma, Lawrence Carin
Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, PMLR 130:100-108, 2021.

Abstract

Naively trained neural networks tend to experience catastrophic forgetting in sequential task settings, where data from previous tasks are unavailable. A number of methods, using various model expansion strategies, have been proposed recently as possible solutions. However, determining how much to expand the model is left to the practitioner, and often a constant schedule is chosen for simplicity, regardless of how complex the incoming task is. Instead, we propose a principled Bayesian nonparametric approach based on the Indian Buffet Process (IBP) prior, letting the data determine how much to expand the model complexity. We pair this with a factorization of the neural network’s weight matrices. Such an approach allows us to scale the number of factors of each weight matrix to the complexity of the task, while the IBP prior encourages sparse weight factor selection and factor reuse, promoting positive knowledge transfer between tasks. We demonstrate the effectiveness of our method on a number of continual learning benchmarks and analyze how weight factors are allocated and reused throughout the training.

Cite this Paper

BibTeX
@InProceedings{pmlr-v130-mehta21a, title = { Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors }, author = {Mehta, Nikhil and Liang, Kevin and Kumar Verma, Vinay and Carin Duke, Lawrence}, booktitle = {Proceedings of The 24th International Conference on Artificial Intelligence and Statistics}, pages = {100--108}, year = {2021}, editor = {Banerjee, Arindam and Fukumizu, Kenji}, volume = {130}, series = {Proceedings of Machine Learning Research}, month = {13--15 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v130/mehta21a/mehta21a.pdf}, url = {https://proceedings.mlr.press/v130/mehta21a.html}, abstract = { Naively trained neural networks tend to experience catastrophic forgetting in sequential task settings, where data from previous tasks are unavailable. A number of methods, using various model expansion strategies, have been proposed recently as possible solutions. However, determining how much to expand the model is left to the practitioner, and often a constant schedule is chosen for simplicity, regardless of how complex the incoming task is. Instead, we propose a principled Bayesian nonparametric approach based on the Indian Buffet Process (IBP) prior, letting the data determine how much to expand the model complexity. We pair this with a factorization of the neural network’s weight matrices. Such an approach allows us to scale the number of factors of each weight matrix to the complexity of the task, while the IBP prior encourages sparse weight factor selection and factor reuse, promoting positive knowledge transfer between tasks. We demonstrate the effectiveness of our method on a number of continual learning benchmarks and analyze how weight factors are allocated and reused throughout the training. } }
Endnote
%0 Conference Paper %T Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors %A Nikhil Mehta %A Kevin Liang %A Vinay Kumar Verma %A Lawrence Carin %B Proceedings of The 24th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2021 %E Arindam Banerjee %E Kenji Fukumizu %F pmlr-v130-mehta21a %I PMLR %P 100--108 %U https://proceedings.mlr.press/v130/mehta21a.html %V 130 %X Naively trained neural networks tend to experience catastrophic forgetting in sequential task settings, where data from previous tasks are unavailable. A number of methods, using various model expansion strategies, have been proposed recently as possible solutions. However, determining how much to expand the model is left to the practitioner, and often a constant schedule is chosen for simplicity, regardless of how complex the incoming task is. Instead, we propose a principled Bayesian nonparametric approach based on the Indian Buffet Process (IBP) prior, letting the data determine how much to expand the model complexity. We pair this with a factorization of the neural network’s weight matrices. Such an approach allows us to scale the number of factors of each weight matrix to the complexity of the task, while the IBP prior encourages sparse weight factor selection and factor reuse, promoting positive knowledge transfer between tasks. We demonstrate the effectiveness of our method on a number of continual learning benchmarks and analyze how weight factors are allocated and reused throughout the training.
APA
Mehta, N., Liang, K., Kumar Verma, V. & Carin, L.. (2021). Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors . Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 130:100-108 Available from https://proceedings.mlr.press/v130/mehta21a.html.

Related Material