Robust Data Pruning: Uncovering and Overcoming Implicit Bias
- URL: http://arxiv.org/abs/2404.05579v1
- Date: Mon, 8 Apr 2024 14:55:35 GMT
- Title: Robust Data Pruning: Uncovering and Overcoming Implicit Bias
- Authors: Artem Vysogorets, Kartik Ahuja, Julia Kempe,
- Abstract summary: We conduct the first systematic study of this effect and reveal that existing data pruning algorithms can produce highly biased classifiers.
We propose a "fairness-aware" approach to pruning and empirically demonstrate its performance on standard computer vision benchmarks.
- Score: 11.930434318557156
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: In the era of exceptionally data-hungry models, careful selection of the training data is essential to mitigate the extensive costs of deep learning. Data pruning offers a solution by removing redundant or uninformative samples from the dataset, which yields faster convergence and improved neural scaling laws. However, little is known about its impact on classification bias of the trained models. We conduct the first systematic study of this effect and reveal that existing data pruning algorithms can produce highly biased classifiers. At the same time, we argue that random data pruning with appropriate class ratios has potential to improve the worst-class performance. We propose a "fairness-aware" approach to pruning and empirically demonstrate its performance on standard computer vision benchmarks. In sharp contrast to existing algorithms, our proposed method continues improving robustness at a tolerable drop of average performance as we prune more from the datasets. We present theoretical analysis of the classification risk in a mixture of Gaussians to further motivate our algorithm and support our findings.
Related papers
- Restoring balance: principled under/oversampling of data for optimal classification [0.0]
Class imbalance in real-world data poses a common bottleneck for machine learning tasks.
Mitigation strategies, such as under or oversampling the data depending on their abundances, are routinely proposed and tested empirically.
We provide a sharp prediction of the effects of under/oversampling strategies depending on class imbalance, first and second moments of the data, and the metrics of performance considered.
arXiv Detail & Related papers (2024-05-15T17:45:34Z) - PUMA: margin-based data pruning [51.12154122266251]
We focus on data pruning, where some training samples are removed based on the distance to the model classification boundary (i.e., margin)
We propose PUMA, a new data pruning strategy that computes the margin using DeepFool.
We show that PUMA can be used on top of the current state-of-the-art methodology in robustness, and it is able to significantly improve the model performance unlike the existing data pruning strategies.
arXiv Detail & Related papers (2024-05-10T08:02:20Z) - Synthetic Information towards Maximum Posterior Ratio for deep learning
on Imbalanced Data [1.7495515703051119]
We propose a technique for data balancing by generating synthetic data for the minority class.
Our method prioritizes balancing the informative regions by identifying high entropy samples.
Our experimental results on forty-one datasets demonstrate the superior performance of our technique.
arXiv Detail & Related papers (2024-01-05T01:08:26Z) - Towards Accelerated Model Training via Bayesian Data Selection [45.62338106716745]
We propose a more reasonable data selection principle by examining the data's impact on the model's generalization loss.
Recent work has proposed a more reasonable data selection principle by examining the data's impact on the model's generalization loss.
This work solves these problems by leveraging a lightweight Bayesian treatment and incorporating off-the-shelf zero-shot predictors built on large-scale pre-trained models.
arXiv Detail & Related papers (2023-08-21T07:58:15Z) - 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) - Efficient and Robust Classification for Sparse Attacks [34.48667992227529]
We consider perturbations bounded by the $ell$--norm, which have been shown as effective attacks in the domains of image-recognition, natural language processing, and malware-detection.
We propose a novel defense method that consists of "truncation" and "adrial training"
Motivated by the insights we obtain, we extend these components to neural network classifiers.
arXiv Detail & Related papers (2022-01-23T21:18:17Z) - Risk Minimization from Adaptively Collected Data: Guarantees for
Supervised and Policy Learning [57.88785630755165]
Empirical risk minimization (ERM) is the workhorse of machine learning, but its model-agnostic guarantees can fail when we use adaptively collected data.
We study a generic importance sampling weighted ERM algorithm for using adaptively collected data to minimize the average of a loss function over a hypothesis class.
For policy learning, we provide rate-optimal regret guarantees that close an open gap in the existing literature whenever exploration decays to zero.
arXiv Detail & Related papers (2021-06-03T09:50:13Z) - Robust Fairness-aware Learning Under Sample Selection Bias [17.09665420515772]
We propose a framework for robust and fair learning under sample selection bias.
We develop two algorithms to handle sample selection bias when test data is both available and unavailable.
arXiv Detail & Related papers (2021-05-24T23:23:36Z) - Low-Regret Active learning [64.36270166907788]
We develop an online learning algorithm for identifying unlabeled data points that are most informative for training.
At the core of our work is an efficient algorithm for sleeping experts that is tailored to achieve low regret on predictable (easy) instances.
arXiv Detail & Related papers (2021-04-06T22:53:45Z) - Learning from Similarity-Confidence Data [94.94650350944377]
We investigate a novel weakly supervised learning problem of learning from similarity-confidence (Sconf) data.
We propose an unbiased estimator of the classification risk that can be calculated from only Sconf data and show that the estimation error bound achieves the optimal convergence rate.
arXiv Detail & Related papers (2021-02-13T07:31:16Z) - Negative Data Augmentation [127.28042046152954]
We show that negative data augmentation samples provide information on the support of the data distribution.
We introduce a new GAN training objective where we use NDA as an additional source of synthetic data for the discriminator.
Empirically, models trained with our method achieve improved conditional/unconditional image generation along with improved anomaly detection capabilities.
arXiv Detail & Related papers (2021-02-09T20:28:35Z)
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.