Continuous Meta-Learning without Tasks

By James Harrison, Apoorva Sharma, Chelsea Finn, Marco Pavone. 2019

84 citations recorded by the source corpus.

Meta-learning is a promising strategy for learning to efficiently learn within new tasks, using data gathered from a distribution of tasks. However, the meta-learning literature thus far has focused on the task segmented setting, where at train-time, offline data is assumed to be split according to the underlying task, and at test-time, the algorithms are optimized to learn in a single task. In this work, we enable the application of generic meta-learning algorithms to settings where this task segmentation is unavailable, such as continual online learning with a time-varying task. We present meta-learning via online changepoint analysis (MOCA), an approach which augments a meta-learning algorithm with a differentiable Bayesian changepoint detection scheme. The framework allows both training and testing directly on time series data without segmenting it into discrete tasks. We demonstrate the utility of this approach on a nonlinear meta-regression benchmark as well as two meta-image-classification benchmarks.

Subfields: Artificial Intelligence, 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 Meta & Continual Learning

Citations in the corpus

Continual Learning via Local Module Composition (2021, background): “Continual-meta learning focuses on fast learning and remembering [25, 35, 33, 41], often emphasising the online performance on OOD tasks [14].”
Gradient-based Editing of Memory Examples for Online Task-free Continual Learning (2020, background): “Due to its broader applicability to realworld data-streams, a number of algorithms have been adapted to the task-free setup [2, 14, 15, 22, 44].”
Incremental Meta-Learning via Indirect Discriminant Alignment (2020, background): “[15, 11] on the other hand, tackled a continual/online meta-learning setup where an explicit delineation between different tasks is not available, whereas in our experimental setup we are primarily trying to solve new classification tasks with clear task boundaries.”
Meta-learning without data via Wasserstein distributionally-robust model fusion (2022, methodology): “…et al., 2020] addresses the problem of fast online adaptation to new tasks while maintaining acquired knowledge on previously learned tasks; MOCA [Harrison et al., 2020] utilizes context data from previous tasks to improve future sequential prediction without knowing when the latent task changes.”
Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning (2020, methodology): “Harrison et al. [23] propose a new CML framework and accompanying model (MOCA).”
Meta Learning on a Sequence of Imbalanced Domains with Difficulty Awareness (2021, unknown): “MOCA [26] works in online learning and learns the experiences from previous data to improve sequential prediction.”
Adaptive Risk Minimization: A Meta-Learning Approach for Tackling Group Shift (2020, methodology): “Some meta-learning methods adapt using both labeled and unlabeled data, such as [53, 71, 46, 2, 39, 4], or consider the setting where task groupings are not known, such as [23, 30, 24, 59], though these works do not focus on the same setting of distribution shift.”
In the Wild: From ML Models to Pragmatic ML Systems (2020, background): “Several setups have been proposed over the years [27, 22, 36, 15, 2, 37, 1] in order to evaluate systems’ abilities to learn continuously and primarily focus on catastrophic forgetting, a phenomenon where models drastically lose accuracy on old tasks when trained on new tasks.”
A CCELERATING O NLINE R EINFORCEMENT L EARNING VIA M ODEL -B ASED M ETA -L EARNING (2021, methodology): “Recent methods do support meta-learning without well-deﬁned task boundaries, but they have only been shown to work on supervised meta-training problems (Harrison et al., 2020; Nagabandi et al., 2019b).”
Deep Reinforcement Learning amidst Continual Structured Non-Stationarity (2021, background): “Previous work within sequential decision making in changing environments (Da Silva et al., 2006; Hadoux et al., 2014; Banerjee et al., 2017) and meta-learning within changing data streams (Harrison et al., 2019) may enable a version of LILAC capable of handling unobserved changepoints.”
Learning to Learn and Remember Super Long Multi-Domain Task Sequence (2022, methodology): “MOCA [26] focuses on meta-learning in online learning, i.e., utilizing more context from previous data to improve future sequential prediction;they are entirely different from SDML. CAVIA [82] uses a separate context vector for fast task adaptation, while SDML focuses on domain-level task…”
Variable-Shot Adaptation for Online Meta-Learning (2020, background): “Finally, our work considers a problem setting that is distinct from works that a study online or continual learning in the inner loop of meta-learning (Al-Shedivat et al., 2017; Nagabandi et al., 2018; Javed & White, 2019; Denevi et al., 2019; He et al., 2019; Harrison et al., 2019).”
Deep Reinforcement Learning amidst Lifelong Non-Stationarity (2020, background): “…making in changing environments (Da Silva et al., 2006; Hadoux et al., 2014; Banerjee et al., 2017) and meta-learning within changing data streams (Harrison et al., 2019) may enable a version of LILAC capable of handling unobserved changepoints, and this setting is likely a fruitful direction for…”
On the Soft-Subnetwork for Few-shot Class Incremental Learning (2022, methodology): “According to the access to task identity during the training and inference phase, the research is often categorized into specific problems, such as task-incremental [34, 8], class-incremental [4, 25, 27, 37, 23, 3, 19, 46], and task-free continual learning [1, 21, 35, 11].”
Wandering Within a World: Online Contextualized Few-Shot Learning (2020, background): “FSL[45]MOCA[15] OnlineMixture[21]OnlineMeta[20] ContinualFSL*[4]OSAKA*[5] OC-FSL(Ours) * denotes concurrent work.”