paper-with-me

홈 › Papers

Be Persistent: Towards a Unified Solution for Mitigating Shortcuts in Deep Learning

2024-02-17 · Hadi M. Dolatabadi, Sarah M. Erfani, Christopher Leckie

Deep neural networks (DNNs) are vulnerable to shortcut learning: rather than learning the intended task, they tend to draw inconclusive relationships between their inputs and outputs. Shortcut learning is ubiquitous among many failure cases of neural networks, and traces of this phenomenon can be seen in their generalizability issues, domain shift, adversarial vulnerability, and even bias towards majority groups. In this paper, we argue that this commonality in the cause of various DNN issues creates a significant opportunity that should be leveraged to find a unified solution for shortcut learning. To this end, we outline the recent advances in topological data analysis (TDA), and persistent homology (PH) in particular, to sketch a unified roadmap for detecting shortcuts in deep learning. We demonstrate our arguments by investigating the topological features of computational graphs in DNNs using two cases of unlearnable examples and bias in decision-making as our test studies. Our analysis of these two failure cases of DNNs reveals that finding a unified solution for shortcut learning in DNNs is not out of reach, and TDA can play a significant role in forming such a framework.

📄 PDF Abstract BibTeX arXiv:2402.11237

Code (0)

등록된 구현이 없습니다.

Tasks

Decision MakingTopological Data Analysis

Similar Papers 제목 키워드 기반

A Whac-A-Mole Dilemma: Shortcuts Come in Multiples Where Mitigating One Amplifies Others

2022-12-09 · CVPR 2023 1 · Zhiheng Li, Ivan Evtimov, Albert Gordo, Caner Hazirbas 외

Machine learning models have been found to learn shortcuts -- unintended decision rules that are unable to generalize -- undermining models' reliability. Previous works address this problem under the tenuous assumption t…

Domain GeneralizationImage ClassificationOut-of-Distribution Generalization

What do neural networks learn in image classification? A frequency shortcut perspective

2023-07-19 · ICCV 2023 1 · Shunxin Wang, Raymond Veldhuis, Christoph Brune, Nicola Strisciuglio

Frequency analysis is useful for understanding the mechanisms of representation learning in neural networks (NNs). Most research in this area focuses on the learning dynamics of NNs for regression tasks, while little for…

Data Augmentationimage-classificationImage ClassificationRepresentation Learning

A Survey on Measuring and Mitigating Reasoning Shortcuts in Machine Reading Comprehension

2022-09-05 · Xanh Ho, Johannes Mario Meissner, Saku Sugawara, Akiko Aizawa

The issue of shortcut learning is widely known in NLP and has been an important research focus in recent years. Unintended correlations in the data enable models to easily solve tasks that were meant to exhibit advanced …

Machine Reading ComprehensionReading Comprehension

Gender Bias in Machine Translation

2021-04-13 · Beatrice Savoldi, Marco Gaido, Luisa Bentivogli, Matteo Negri 외

Machine translation (MT) technology has facilitated our daily tasks by providing accessible shortcuts for gathering, elaborating and communicating information. However, it can suffer from biases that harm users and socie…

Machine TranslationTranslation

Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models

2021-10-14 · Findings (NAACL) 2022 7 · Tianlu Wang, Rohit Sridhar, Diyi Yang, Xuezhi Wang

Recently, NLP models have achieved remarkable progress across a variety of tasks; however, they have also been criticized for being not robust. Many robustness problems can be attributed to models exploiting spurious cor…