Coresets via Bilevel Optimization for Continual Learning and Streaming

By Zalán Borsos, Mojm'ir Mutn'y, Andreas Krause. 2020

307 citations recorded by the source corpus.

Coresets are small data summaries that are sufficient for model training. They can be maintained online, enabling efficient handling of large data streams under resource constraints. However, existing constructions are limited to simple models such as k-means and logistic regression. In this work, we propose a novel coreset construction via cardinality-constrained bilevel optimization. We show how our framework can efficiently generate coresets for deep neural networks, and demonstrate its empirical benefits in continual learning and in streaming settings.

Subfields: Artificial Intelligence, Computational Mechanics, Optimization & Training, 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 Computational Mechanics

In Optimization & Training

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “…by maximizing such as the sample diversity of parameter gradients [21], performance of corresponding tasks with cardinality constraints [132], mini-batch gradient similarity and cross-batch gradient diversity [133], ability of optimizing latent decision boundaries [134], diversity of…”
The Ideal Continual Learner: An Agent That Never Forgets (2023, background): “Note that bilevel programming has been used for continual learning (Borsos et al., 2020), but the use is mainly for selecting which data samples to store in the memory.”
Information-theoretic Online Memory Selection for Continual Learning (2022, neutral): “…sampling with the output-space Hessian (WRS), class-balanced reservoir sampling (CBRS) (Chrysakis & Moens, 2020), greedy gradient sample selection (GSS) (Aljundi et al., 2019b), feature-space clustering (FSS-Clust) (Aljundi et al., 2019b), coreset (Borsos et al., 2020), InfoGS, and InfoRS.”
Constrained Few-shot Class-incremental Learning (2022, methodology): “To address the catastrophic forgetting issue in CIL, rehearsal methods store some old training data, and replay them while learning novel classes during the incremental stage [1, 3, 31, 32, 36, 39].”
Online Coreset Selection for Rehearsal-based Continual Learning (2021, critical): “[5] propose a bilevel optimization framework with cardinality constraints for coreset selection.”
New Insights on Reducing Abrupt Representation Change in Online Continual Learning (2021, methodology): “Several works (Borsos et al., 2020; Chaudhry et al., 2019; Aljundi et al., 2019b) propose solutions to select which samples should be stored, or retrieved for replay (Aljundi et al.”
Rainbow Memory: Continual Learning with a Memory of Diverse Samples (2021, methodology): “Also, Borsos et al . propose a coreset generation method for the representative memory using cardinality-constrained bi-level optimization [3]. and Cong et al . propose a GAN based memory which they can perturb styles of remembered samples for incremental learning [8].”
Posterior Meta-Replay for Continual Learning (2021, neutral): “[6] Z.”
The Effectiveness of Memory Replay in Large Scale Continual Learning (2020, neutral): “…new update rules using replay samples (Lopez-Paz & Ranzato, 2017; Chaudhry et al., 2018), while others focus on new sampling strategies to ﬁll the replay buffer (Rebufﬁ et al., 2017; Chaudhry et al., 2020; Aljundi et al., 2019; Hayes et al., 2019; Chrysakis & Moens, 2020; Borsos et al., 2020).”
SVP-CF: Selection via Proxy for Collaborative Filtering Data (2021, background): “More recent works treat coreset selection as a bi-level optimization problem (Borsos et al., 2020; Krause et al., 2021) and directly optimize for the best possible subset for the downstream task.”
Lifelong Event Detection with Knowledge Transfer (2021, methodology): “The Learning without For-getting method (Li and Hoiem, 2018) Other work uses bi-level optimization to select better exemplar instances (Liu et al., 2020; Borsos et al., 2020).”
Reducing Representation Drift in Online Continual Learning (2021, background): “Various works focus on studying which samples to store (Borsos et al., 2020; Aljundi et al., 2019) or which samples to replay when receiving new data (Caccia et al.”
Retrospective Adversarial Replay for Continual Learning (2022, background): “This category also includes memory-based approaches for constrained parameter updates on new samples to minimize interference with previous tasks [9, 37] and for replaying them along with the new task’s data [2, 3, 5, 6, 10, 13, 38, 41, 43, 44, 47, 48, 59, 55, 65].”
Dataset Distillation with Infinitely Wide Convolutional Networks (2021, methodology): “L G ] 2 7 O ct 2 effectiveness of replay methods in continual learning [Borsos et al., 2020] and helping to accelerate neural architecture search [Zhao et al., 2021, Zhao and Bilen, 2021].”
How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition (2022, background): “One additional approach determines the optimal points to store in memory using bilevel optimization (Borsos et al., 2020).”
Dataset Meta-Learning from Kernel Ridge-Regression (2020, methodology): “While there are simple baselines based on random sampling, greedy sampling and clustering mechanism, one notable work is Borsos et al. (2020), utilizing KRR for dataset subselection.”