Large-scale Robust Deep AUC Maximization: A New Surrogate Loss and Empirical Studies on Medical Image Classification (2020-12-06T00:00:00.000000Z)

TL;DR

This work proposes a new margin-based min-max surrogate loss function for the AUC score that is more robust than the commonly used AUC square loss, while enjoying the same advantage in terms of large-scale stochastic optimization.

Abstract

Deep AUC Maximization (DAM) is a new paradigm for learning a deep neural network by maximizing the AUC score of the model on a dataset. Most previous works of AUC maximization focus on the perspective of optimization by designing efficient stochastic algorithms, and studies on generalization performance of large-scale DAM on difficult tasks are missing. In this work, we aim to make DAM more practical for interesting real-world applications (e.g., medical image classification). First, we propose a new margin-based min-max surrogate loss function for the AUC score (named as the AUC min-max-margin loss or simply AUC margin loss for short). It is more robust than the commonly used AUC square loss, while enjoying the same advantage in terms of large-scale stochastic optimization. Second, we conduct extensive empirical studies of our DAM method on four difficult medical image classification tasks, namely (i) classification of chest x-ray images for identifying many threatening diseases, (ii) classification of images of skin lesions for identifying melanoma, (iii) classification of mammogram for breast cancer screening, and (iv) classification of microscopic images for identifying tumor tissue. Our studies demonstrate that the proposed DAM method improves the performance of optimizing cross-entropy loss by a large margin, and also achieves better performance than optimizing the existing AUC square loss on these medical image classification tasks. Specifically, our DAM method has achieved the 1st place on Stanford CheXpert competition on Aug. 31, 2020. To the best of our knowledge, this is the first work that makes DAM succeed on large-scale medical image datasets. We also conduct extensive ablation studies to demonstrate the advantages of the new AUC margin loss over the AUC square loss on benchmark datasets. The proposed method is implemented in our open-sourced library LibAUC (www.libauc.org) whose github address is https://github.com/Optimization-AI/LibAUC.

References41 items

Federated Deep AUC Maximization for Heterogeneous Data with a Constant Communication Complexity

Identifying Melanoma Images using EfficientNet Ensemble: Winning Solution to the SIIM-ISIC Melanoma Classification Challenge

A patient-centric dataset of images and metadata for identifying melanomas using clinical context

Fast Objective & Duality Gap Convergence for Non-Convex Strongly-Concave Min-Max Problems with PL Condition

Weakly Supervised Lesion Localization With Probabilistic-CAM Pooling

Communication-Efficient Distributed Stochastic AUC Maximization with Deep Neural Networks

Interpreting Chest X-rays via CNNs that Exploit Hierarchical Disease Dependencies and Uncertainty Labels

Decoupling Representation and Classifier for Long-Tailed Recognition

Stochastic AUC Maximization with Deep Neural Networks

EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

Deep Neural Networks Improve Radiologists’ Performance in Breast Cancer Screening

CheXpert: A Large Chest Radiograph Dataset with Uncertainty Labels and Expert Comparison

Melanoma

Stochastic Proximal Algorithms for AUC Maximization

Fast Stochastic AUC Maximization with O(1/n)-Convergence Rate

Rotation Equivariant CNNs for Digital Pathology

Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer

Focal Loss for Dense Object Detection

Dermatologist-level classification of skin cancer with deep neural networks

Stochastic Online AUC Maximization

Densely Connected Convolutional Networks

Skin lesion analysis toward melanoma detection: A challenge at the 2017 International symposium on biomedical imaging (ISBI), hosted by the international skin imaging collaboration (ISIC)

Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning

Deep Residual Learning for Image Recognition

Cyclical Learning Rates for Training Neural Networks

Very Deep Convolutional Networks for Large-Scale Image Recognition

One-Pass AUC Optimization

On the Consistency of AUC Pairwise Optimization

An Analysis of Single-Layer Networks in Unsupervised Feature Learning

Online AUC Maximization

Asirra: a CAPTCHA that exploits interest-aligned manual image categorization

Ranking and empirical minimization of U-statistics

A support vector method for multivariate performance measures

Optimising area under the ROC curve using gradient descent

AUC Optimization vs. Error Rate Minimization

The meaning and use of the area under a receiver operating characteristic (ROC) curve.

Triple stratified kfold with tfrecords

Maximizing AUC with Deep Learning for Classification of Imbalanced Mammogram Datasets

Learning Multiple Layers of Features from Tiny Images

THE DIGITAL DATABASE FOR SCREENING MAMMOGRAPHY

Current Status of the Digital Database for Screening Mammography