Architect, Regularize and Replay (ARR): a Flexible Hybrid Approach for Continual Learning
Preprint
Abstract
In recent years we have witnessed a renewed interest in machine learning methodologies, especially for deep representation learning, that could overcome basic i.i.d. assumptions and tackle non-stationary environments subject to various distributional shifts or sample selection biases. Within this context, several computational approaches based on architectural priors, regularizers and replay policies have been proposed with different degrees of success depending on the specific scenario in which they were developed and assessed. However, designing comprehensive hybrid solutions that can flexibly and generally be applied with tunable efficiency-effectiveness trade-offs still seems a distant goal. In this paper, we propose "Architect, Regularize and Replay" (ARR), an hybrid generalization of the renowned AR1 algorithm and its variants, that can achieve state-of-the-art results in classic scenarios (e.g. class-incremental learning) but also generalize to arbitrary data streams generated from real-world datasets such as CIFAR-100, CORe50 and ImageNet-1000.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- Continual pre-training mitigates forgetting in language and visionNeural Networks, 2024
- Generative negative replay for continual learningNeural Networks, 2023
- Practical Recommendations for Replay-Based Continual Learning Methods2022
- Ex-Model: Continual Learning from a Stream of Trained Models2022
- CVPR 2020 continual learning in computer vision competition: Approaches, results, current challenges and future directionsArtificial Intelligence, 2022
- Latent Replay for Real-Time Continual Learning2020
- Continual Reinforcement Learning in 3D Non-stationary Environments2020
- Continual learning for robotics: Definition, framework, learning strategies, opportunities and challengesInformation Fusion, 2020
- Continuous learning in single-incremental-task scenariosNeural Networks, 2019
- Learning without ForgettingIEEE Transactions on Pattern Analysis and Machine Intelligence, 2017