Improving Online Continual Learning Performance and Stability with Temporal Ensembles

• URL: http://arxiv.org/abs/2306.16817v2
• Date: Mon, 3 Jul 2023 14:55:43 GMT
• Title: Improving Online Continual Learning Performance and Stability with Temporal Ensembles
• Authors: Albin Soutif--Cormerais, Antonio Carta, Joost Van de Weijer
• Abstract summary: We study the effect of model ensembling as a way to improve performance and stability in online continual learning. We use a lightweight temporal ensemble that computes the exponential moving average of the weights (EMA) at test time.
• Score: 30.869268130955145
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Neural networks are very effective when trained on large datasets for a large number of iterations. However, when they are trained on non-stationary streams of data and in an online fashion, their performance is reduced (1) by the online setup, which limits the availability of data, (2) due to catastrophic forgetting because of the non-stationary nature of the data. Furthermore, several recent works (Caccia et al., 2022; Lange et al., 2023) arXiv:2205.13452 showed that replay methods used in continual learning suffer from the stability gap, encountered when evaluating the model continually (rather than only on task boundaries). In this article, we study the effect of model ensembling as a way to improve performance and stability in online continual learning. We notice that naively ensembling models coming from a variety of training tasks increases the performance in online continual learning considerably. Starting from this observation, and drawing inspirations from semi-supervised learning ensembling methods, we use a lightweight temporal ensemble that computes the exponential moving average of the weights (EMA) at test time, and show that it can drastically increase the performance and stability when used in combination with several methods from the literature.

Related papers

• Towards An Online Incremental Approach to Predict Students Performance [0.8287206589886879]
  We propose a memory-based online incremental learning approach for updating an online classifier. Our approach achieves a notable improvement in model accuracy, with an enhancement of nearly 10% compared to the current state-of-the-art.
  arXiv  Detail & Related papers  (2024-05-03T17:13:26Z)
• Small Dataset, Big Gains: Enhancing Reinforcement Learning by Offline Pre-Training with Model Based Augmentation [59.899714450049494]
  offline pre-training can produce sub-optimal policies and lead to degraded online reinforcement learning performance. We propose a model-based data augmentation strategy to maximize the benefits of offline reinforcement learning pre-training and reduce the scale of data needed to be effective.
  arXiv  Detail & Related papers  (2023-12-15T14:49:41Z)
• VERSE: Virtual-Gradient Aware Streaming Lifelong Learning with Anytime Inference [36.61783715563126]
  Streaming lifelong learning is a challenging setting of lifelong learning with the goal of continuous learning without forgetting. We introduce a novel approach to lifelong learning, which is streaming (observes each training example only once) We propose a novel emphvirtual gradients based approach for continual representation learning which adapts to each new example while also generalizing well on past data to prevent catastrophic forgetting.
  arXiv  Detail & Related papers  (2023-09-15T07:54:49Z)
• Robust Learning with Progressive Data Expansion Against Spurious Correlation [65.83104529677234]
  We study the learning process of a two-layer nonlinear convolutional neural network in the presence of spurious features. Our analysis suggests that imbalanced data groups and easily learnable spurious features can lead to the dominance of spurious features during the learning process. We propose a new training algorithm called PDE that efficiently enhances the model's robustness for a better worst-group performance.
  arXiv  Detail & Related papers  (2023-06-08T05:44:06Z)
• Model-Based Reinforcement Learning with Multi-Task Offline Pretraining [59.82457030180094]
  We present a model-based RL method that learns to transfer potentially useful dynamics and action demonstrations from offline data to a novel task. The main idea is to use the world models not only as simulators for behavior learning but also as tools to measure the task relevance. We demonstrate the advantages of our approach compared with the state-of-the-art methods in Meta-World and DeepMind Control Suite.
  arXiv  Detail & Related papers  (2023-06-06T02:24:41Z)
• Dissecting Continual Learning a Structural and Data Analysis [0.0]
  Continual Learning is a field dedicated to devise algorithms able to achieve lifelong learning. Deep learning methods can attain impressive results when the data modeled does not undergo a considerable distributional shift in subsequent learning sessions. When we expose such systems to this incremental setting, performance drop very quickly.
  arXiv  Detail & Related papers  (2023-01-03T10:37:11Z)
• Towards Robust Dataset Learning [90.2590325441068]
  We propose a principled, tri-level optimization to formulate the robust dataset learning problem. Under an abstraction model that characterizes robust vs. non-robust features, the proposed method provably learns a robust dataset.
  arXiv  Detail & Related papers  (2022-11-19T17:06:10Z)
• Passive learning to address nonstationarity in virtual flow metering applications [0.0]
  This paper explores how learning methods can be applied to sustain the prediction accuracy of steady-state virtual flow meters. Two passive learning methods, periodic batch learning and online learning, are applied with varying calibration frequency to train virtual flow meters. The results are two-fold: first, in the presence of frequently arriving measurements, frequent model updating sustains an excellent prediction performance over time; second, in the presence of intermittent and infrequently arriving measurements, frequent updating is essential to increase the performance accuracy.
  arXiv  Detail & Related papers  (2022-02-07T14:42:00Z)
• Online Continual Learning with Natural Distribution Shifts: An Empirical Study with Visual Data [101.6195176510611]
  "Online" continual learning enables evaluating both information retention and online learning efficacy. In online continual learning, each incoming small batch of data is first used for testing and then added to the training set, making the problem truly online. We introduce a new benchmark for online continual visual learning that exhibits large scale and natural distribution shifts.
  arXiv  Detail & Related papers  (2021-08-20T06:17:20Z)
• Tracking Performance of Online Stochastic Learners [57.14673504239551]
  Online algorithms are popular in large-scale learning settings due to their ability to compute updates on the fly, without the need to store and process data in large batches. When a constant step-size is used, these algorithms also have the ability to adapt to drifts in problem parameters, such as data or model properties, and track the optimal solution with reasonable accuracy. We establish a link between steady-state performance derived under stationarity assumptions and the tracking performance of online learners under random walk models.
  arXiv  Detail & Related papers  (2020-04-04T14:16:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.