Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors

By Nikhil Mehta, Kevin J Liang, Vinay Kumar Verma, Lawrence Carin. 2021

Subfields: Probabilistic & Bayesian Methods, 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 Probabilistic & Bayesian Methods

1. Piyush Rai
2. Liyuan Wang
3. Kevin J Liang
4. Lawrence Carin
5. Jun Zhu
6. Vinay Kumar Verma
7. Panos P. Markopoulos
8. Nikhil Mehta

In Meta & Continual Learning

1. Xingxing Zhang
2. Liyuan Wang
3. Hang Su
4. Jun Zhu
5. Piyush Rai
6. Vinay Kumar Verma
7. Ashley Prater-Bennette
8. Kevin J Liang

Citations in the corpus

• A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, methodology): “The function of task-identity prediction can also be learned from incremental tasks, using other continual learningstrategiestomitigatecatastrophicforgetting[78],[223], [233], [241].”
• Incremental Task Learning with Incremental Rank Updates (2022, methodology): “A factorization-based approach was proposed in [22] that performs automatic rank selection per task for variational inference using Indian Buffet process.”
• Efficient Feature Transformations for Discriminative and Generative Continual Learning (2021, methodology): “Alternatively, some recent works leverage Bayesian nonparametrics to let the data dictate expansion [32, 21, 23], but the benchmarks considered in Bayesian methods have been limited to MNIST and CIFAR-10.”
• A Metaverse: Taxonomy, Components, Applications, and Open Challenges (2022, unknown): “To solve the catastrophic forgetting problem in a sequential task where data from previous tasks are not available, Mehta et al. [319] proposed a principled Bayesian nonparametric approach, the Indian Buffet Process (IBP), which determines how much the data scales to model complexity.”