Change is Hard: A Closer Look at Subpopulation Shift
- URL: http://arxiv.org/abs/2302.12254v3
- Date: Thu, 17 Aug 2023 16:15:28 GMT
- Title: Change is Hard: A Closer Look at Subpopulation Shift
- Authors: Yuzhe Yang, Haoran Zhang, Dina Katabi, Marzyeh Ghassemi
- Abstract summary: We propose a unified framework that dissects and explains common shifts in subgroups.
We then establish a benchmark of 20 state-of-the-art algorithms evaluated on 12 real-world datasets in vision, language, and healthcare domains.
- Score: 48.0369745740936
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Machine learning models often perform poorly on subgroups that are
underrepresented in the training data. Yet, little is understood on the
variation in mechanisms that cause subpopulation shifts, and how algorithms
generalize across such diverse shifts at scale. In this work, we provide a
fine-grained analysis of subpopulation shift. We first propose a unified
framework that dissects and explains common shifts in subgroups. We then
establish a comprehensive benchmark of 20 state-of-the-art algorithms evaluated
on 12 real-world datasets in vision, language, and healthcare domains. With
results obtained from training over 10,000 models, we reveal intriguing
observations for future progress in this space. First, existing algorithms only
improve subgroup robustness over certain types of shifts but not others.
Moreover, while current algorithms rely on group-annotated validation data for
model selection, we find that a simple selection criterion based on worst-class
accuracy is surprisingly effective even without any group information. Finally,
unlike existing works that solely aim to improve worst-group accuracy (WGA), we
demonstrate the fundamental tradeoff between WGA and other important metrics,
highlighting the need to carefully choose testing metrics. Code and data are
available at: https://github.com/YyzHarry/SubpopBench.
Related papers
- OOD-Chameleon: Is Algorithm Selection for OOD Generalization Learnable? [18.801143204410913]
We formalize the task of algorithm selection for OOD generalization and investigate whether it could be approached by learning.
We propose a solution, dubbed OOD-Chameleon that treats the task as a supervised classification over candidate algorithms.
We train the model to predict the relative performance of algorithms given a dataset's characteristics.
arXiv Detail & Related papers (2024-10-03T17:52:42Z) - The Group Robustness is in the Details: Revisiting Finetuning under Spurious Correlations [8.844894807922902]
Modern machine learning models are prone to over-reliance on spurious correlations.
In this paper, we identify surprising and nuanced behavior of finetuned models on worst-group accuracy.
Our results show more nuanced interactions of modern finetuned models with group robustness than was previously known.
arXiv Detail & Related papers (2024-07-19T00:34:03Z) - Mind the GAP: Improving Robustness to Subpopulation Shifts with Group-Aware Priors [46.03963664373476]
We develop a family of group-aware prior (GAP) distributions over neural network parameters that explicitly favor models that generalize well under subpopulation shifts.
We demonstrate that training withGAP yields state-of-the-art performance -- even when only retraining the final layer of a previously trained non-robust model.
arXiv Detail & Related papers (2024-03-14T21:00:26Z) - Parametric Classification for Generalized Category Discovery: A Baseline
Study [70.73212959385387]
Generalized Category Discovery (GCD) aims to discover novel categories in unlabelled datasets using knowledge learned from labelled samples.
We investigate the failure of parametric classifiers, verify the effectiveness of previous design choices when high-quality supervision is available, and identify unreliable pseudo-labels as a key problem.
We propose a simple yet effective parametric classification method that benefits from entropy regularisation, achieves state-of-the-art performance on multiple GCD benchmarks and shows strong robustness to unknown class numbers.
arXiv Detail & Related papers (2022-11-21T18:47:11Z) - Outlier-Robust Group Inference via Gradient Space Clustering [50.87474101594732]
Existing methods can improve the worst-group performance, but they require group annotations, which are often expensive and sometimes infeasible to obtain.
We address the problem of learning group annotations in the presence of outliers by clustering the data in the space of gradients of the model parameters.
We show that data in the gradient space has a simpler structure while preserving information about minority groups and outliers, making it suitable for standard clustering methods like DBSCAN.
arXiv Detail & Related papers (2022-10-13T06:04:43Z) - Towards Group Robustness in the presence of Partial Group Labels [61.33713547766866]
spurious correlations between input samples and the target labels wrongly direct the neural network predictions.
We propose an algorithm that optimize for the worst-off group assignments from a constraint set.
We show improvements in the minority group's performance while preserving overall aggregate accuracy across groups.
arXiv Detail & Related papers (2022-01-10T22:04:48Z) - A Fine-Grained Analysis on Distribution Shift [24.084676204709723]
We introduce a framework that enables fine-grained analysis of various distribution shifts.
We evaluate 19 distinct methods grouped into five categories across both synthetic and real-world datasets.
Our framework can be easily extended to include new methods, shifts, and datasets.
arXiv Detail & Related papers (2021-10-21T17:57:08Z) - Examining and Combating Spurious Features under Distribution Shift [94.31956965507085]
We define and analyze robust and spurious representations using the information-theoretic concept of minimal sufficient statistics.
We prove that even when there is only bias of the input distribution, models can still pick up spurious features from their training data.
Inspired by our analysis, we demonstrate that group DRO can fail when groups do not directly account for various spurious correlations.
arXiv Detail & Related papers (2021-06-14T05:39:09Z) - No Fear of Heterogeneity: Classifier Calibration for Federated Learning
with Non-IID Data [78.69828864672978]
A central challenge in training classification models in the real-world federated system is learning with non-IID data.
We propose a novel and simple algorithm called Virtual Representations (CCVR), which adjusts the classifier using virtual representations sampled from an approximated ssian mixture model.
Experimental results demonstrate that CCVR state-of-the-art performance on popular federated learning benchmarks including CIFAR-10, CIFAR-100, and CINIC-10.
arXiv Detail & Related papers (2021-06-09T12:02:29Z) - SetConv: A New Approach for Learning from Imbalanced Data [29.366843553056594]
We propose a set convolution operation and an episodic training strategy to extract a single representative for each class.
We prove that our proposed algorithm is permutation-invariant despite the order of inputs.
arXiv Detail & Related papers (2021-04-03T22:33:30Z)
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.