Wild-Time: A Benchmark of in-the-Wild Distribution Shift over Time
28. listopadu 2022
Řečníci
Huaxiu Yao
Sprecher:in · 0 Follower:innen
Caroline Choi
Sprecher:in · 0 Follower:innen
Yoonho Lee
Sprecher:in · 0 Follower:innen
Pang Wei Koh
Sprecher:in · 0 Follower:innen
Chelsea Finn
Sprecher:in · 57 Follower:innen
O prezentaci
Distribution shifts occur when the test distribution differs from the training distribution, and can considerably degrade performance of machine learning models deployed in the real world. While recent works have studied robustness to distribution shifts, distribution shifts arising from the passage of time have the additional structure of timestamp metadata. Real-world examples of such shifts are underexplored, and it is unclear whether existing models can leverage trends in past distribution shifts to reliably extrapolate into the future. To address this gap, we curate Wild-Time, a benchmark of 7 datasets that reflect temporal distribution shifts arising in a variety of real-world applications, including drug discovery, patient prognosis, and news classification. On these datasets, we systematically benchmark 13 approaches with various inductive biases. We evaluate methods in domain-generalization, continual learning, self-supervised learning, and ensemble learning, which leverage timestamps to extract the common structure of the distribution shifts. We extend several domain-generalization methods to the temporal distribution shift setting by treating windows of time as different domains. Finally, we propose two evaluation strategies to evaluate model performance under temporal distribution shifts—evaluation with a fixed time split (Eval-Fix) and evaluation with a data stream (Eval-Stream). Eval-Fix, our primary evaluation strategy, aims to provide a simple evaluation protocol for the broader machine learning community, while Eval-Stream serves as a complementary benchmark for continual learning approaches. Our experiments demonstrate that existing methods are limited in tackling temporal distribution shift: across all settings, we observe an average performance drop of 20Distribution shifts occur when the test distribution differs from the training distribution, and can considerably degrade performance of machine learning models deployed in the real world. While recent works have studied robustness to distribution shifts, distribution shifts arising from the passage of time have the additional structure of timestamp metadata. Real-world examples of such shifts are underexplored, and it is unclear whether existing models can leverage trends in past distribution s…
Organizátor
NeurIPS 2022
Konto · 963 Follower:innen
Baví vás formát? Nechte SlidesLive zachytit svou akci!
Profesionální natáčení a streamování po celém světě.
Sdílení
Doporučená videa
Prezentace na podobné téma, kategorii nebo přednášejícího
InsNet: An Efficient, Flexible, and Performant Insertion-based Text Generation Model
Später ansehen
Sidi Lu, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Panel RL Implementation
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%
A Neural Corpus Indexer for Document Retrieval
Später ansehen
Yujing Wang, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Temporal Effective Batch Normalization in Spiking Neural Networks
Später ansehen
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Discovery of Single Independent Latent Variable
Später ansehen
Uri Shaham, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%
Cross Aggregation Transformer for Image Restoration
Später ansehen
Zheng Chen, …
Ewigspeicher-Fortschrittswert: 0 = 0.0%