Reconciling meta-learning and continual learning with online mixtures of tasks

Jerfel, Ghassen; Grant, Erin; Griffiths, Thomas L.; Heller, Katherine

doi:10.48550/arxiv.1812.06080

Cited by 3 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Javed and White [203] proposed a meta-objective to learn representations which are naturally highly sparse and thus effectively mitigate catastrophic forgetting in continual learning. Jerfel et al [204] employed a meta-learner to control the amount of knowledge transfer between tasks and automatically adapt to a new task when a task distribution shift is detected. Rajasegaran et al [205] proposed an Incremental Task-Agnostic Meta-learning (iTAML) method, which adapts meta learning to separate the task-agnostic feature extractor from the task-specific classifier.…”

Section: Other Methodsmentioning

confidence: 99%

Recent Advances of Continual Learning in Computer Vision: An Overview

Qu,

Rahmani,

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

Section: Other Methodsmentioning

confidence: 99%

Recent Advances of Continual Learning in Computer Vision: An Overview

Qu,

Rahmani,

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…MAML aims at finding a set of joined task parameters that can be easily fine-tuned to new test tasks via few gradient descent updates. MAML can also be treated as a Bayesian hierarchical model [10,15,18]. Bayesian MAML [55] combines efficient gradient-based meta-learning with non-parametric variational inference in a principled probabilistic framework.…”

Section: Related Workmentioning

confidence: 99%

Non-Gaussian Gaussian Processes for Few-Shot Regression

Sendera¹,

Tabor²,

Nowak³

et al. 2021

Preprint

View full text Add to dashboard Cite

Gaussian Processes (GPs) have been widely used in machine learning to model distributions over functions, with applications including multi-modal regression, time-series prediction, and few-shot learning. GPs are particularly useful in the last application since they rely on Normal distributions and enable closed-form computation of the posterior probability function. Unfortunately, because the resulting posterior is not flexible enough to capture complex distributions, GPs assume high similarity between subsequent tasks -a requirement rarely met in real-world conditions. In this work, we address this limitation by leveraging the flexibility of Normalizing Flows to modulate the posterior predictive distribution of the GP. This makes the GP posterior locally non-Gaussian, therefore we name our method Non-Gaussian Gaussian Processes (NGGPs). We propose an invertible ODE-based mapping that operates on each component of the random variable vectors and shares the parameters across all of them. We empirically tested the flexibility of NGGPs on various few-shot learning regression datasets, showing that the mapping can incorporate context embedding information to model different noise levels for periodic functions. As a result, our method shares the structure of the problem between subsequent tasks, but the contextualization allows for adaptation to dissimilarities. NGGPs outperform the competing state-of-the-art approaches on a diversified set of benchmarks and applications.

show abstract

“…However, standard meta-learning algorithms operate in batch mode, making them poorly suited for continuously evolving environments. More recently, online meta-learning methods have been proposed with the goal of enabling continual adaptation Jerfel et al, 2018;Yao et al, 2020;Nagabandi et al, 2018;Li & Hospedales, 2020), where a constant stream of data from distinct tasks is used for both adaptation and meta-training. In this scheme, meta-training is used to accelerate how quickly the network can adapt to each new task it sees, and simultaneously use that data from each new task for meta-training.…”

Section: Introductionmentioning

confidence: 99%

Fully Online Meta-Learning Without Task Boundaries

Rajasegaran¹,

Finn²,

Levine³

2022

Preprint

View full text Add to dashboard Cite

While deep networks can learn complex functions such as classifiers, detectors, and trackers, many applications require models that continually adapt to changing input distributions, changing tasks, and changing environmental conditions. Indeed, this ability to continuously accrue knowledge and use past experience to learn new tasks quickly in continual settings is one of the key properties of an intelligent system. For complex and high-dimensional problems, simply updating the model continually with standard learning algorithms such as gradient descent may result in slow adaptation. Meta-learning can provide a powerful tool to accelerate adaptation yet is conventionally studied in batch settings. In this paper, we study how metalearning can be applied to tackle online problems of this nature, simultaneously adapting to changing tasks and input distributions and meta-training the model in order to adapt more quickly in the future. Extending meta-learning into the online setting presents its own challenges, and although several prior methods have studied related problems, they generally require a discrete notion of tasks, with known ground-truth task boundaries. Such methods typically adapt to each task in sequence, resetting the model between tasks, rather than adapting continuously across tasks. In many real-world settings, such discrete boundaries are unavailable, and may not even exist. To address these settings, we propose a Fully Online Meta-Learning (FOML) algorithm, which does not require any ground truth knowledge about the task boundaries and stays fully online without resetting back to pre-trained weights. Our experiments show that FOML was able to learn new tasks faster than the state-of-the-art online learning methods on Rainbow-MNIST, CIFAR100 and CELEBA datasets.

show abstract

Reconciling meta-learning and continual learning with online mixtures of tasks

Cited by 3 publications

References 31 publications

Recent Advances of Continual Learning in Computer Vision: An Overview

Recent Advances of Continual Learning in Computer Vision: An Overview

Non-Gaussian Gaussian Processes for Few-Shot Regression

Fully Online Meta-Learning Without Task Boundaries

Contact Info

Product

Resources

About