Distributionally Robust Deep Learning using Hardness Weighted Sampling

Abstract

Limiting failures of machine learning systems is of paramount importance for safety-critical applications. In order to improve the robustness of machine learning systems, Distributionally Robust Optimization (DRO) has been proposed as a generalization of Empirical Risk Minimization (ERM). However, its use in deep learning has been severely restricted due to the relative inefficiency of the optimizers available for DRO in comparison to the wide-spread variants of Stochastic Gradient Descent (SGD) optimizers for ERM. We propose SGD with hardness weighted sampling, a principled and efficient optimization method for DRO in machine learning that is particularly suited in the context of deep learning. Similar to a hard example mining strategy in practice, the proposed algorithm is straightforward to implement and computationally as efficient as SGD-based optimizers used for deep learning, requiring minimal overhead computation. In contrast to typical ad hoc hard mining approaches, we prove the convergence of our DRO algorithm for over-parameterized deep learning networks with ReLU activation and a finite number of layers and parameters. Our experiments on fetal brain 3D MRI segmentation and brain tumor segmentation in MRI demonstrate the feasibility and the usefulness of our approach. Using our hardness weighted sampling for training a state-of-the-art deep learning pipeline leads to improved robustness to anatomical variabilities in automatic fetal brain 3D MRI segmentation using deep learning and to improved robustness to the image protocol variations in brain tumor segmentation. Our code is available at https://github.com/LucasFidon/HardnessWeightedSampler.

arXiv commentAccepted for publication at the Journal of Machine Learning for Biomedical Imaging (MELBA) https://www.melba-journal.org/papers/2022:019.html

References in corpus (21)

• Instance Normalization: The Missing Ingredient for Fast Stylization
• Deeply-Supervised Nets
• A Convergence Theory for Deep Learning via Over-Parameterization
• Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
• Certifying Some Distributional Robustness with Principled Adversarial Training
• On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems
• Online Batch Selection for Faster Training of Neural Networks
• An automatic multi-tissue human fetal brain segmentation benchmark using the Fetal Tissue Annotation Dataset
• Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples
• What is the Effect of Importance Weighting in Deep Learning?
• On the Convergence Rate of Training Recurrent Neural Networks
• Does Distributionally Robust Supervised Learning Give Robust Classifiers?
• Generalised Wasserstein Dice Score for Imbalanced Multi-class Segmentation using Holistic Convolutional Networks
• Sampling Matters in Deep Embedding Learning
• Fetal Brain Tissue Annotation and Segmentation Challenge Results
• Generalization Error Bounds of Gradient Descent for Learning Over-parameterized Deep ReLU Networks
• An Improved Analysis of Training Over-parameterized Deep Neural Networks
• Label-set Loss Functions for Partial Supervision: Application to Fetal Brain 3D MRI Parcellation
• Distributionally Robust Segmentation of Abnormal Fetal Brain 3D MRI
• MONAIfbs: MONAI-based fetal brain MRI deep learning segmentation
• Comparative study of deep learning methods for the automatic segmentation of lung, lesion and lesion type in CT scans of COVID-19 patients

Cited by in corpus (2)

• A Dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation
• Distributionally Robust Segmentation of Abnormal Fetal Brain 3D MRI