Timezone: »
Deep neural networks (DNNs) are notorious for making more mistakes for the classes that have substantially fewer samples than the others during training. Such class imbalance is ubiquitous in clinical applications and very crucial to handle because the classes with fewer samples most often correspond to critical cases (e.g., cancer) where misclassifications can have severe consequences.Not to miss such cases, binary classifiers need to be operated at high True Positive Rates (TPRs) by setting a higher threshold, but this comes at the cost of very high False Positive Rates (FPRs) for problems with class imbalance. Existing methods for learning under class imbalance most often do not take this into account. We argue that prediction accuracy should be improved by emphasizing the reduction of FPRs at high TPRs for problems where misclassification of the positive, i.e. critical, class samples are associated with higher cost.To this end, we pose the training of a DNN for binary classification as a constrained optimization problem and introduce a novel constraint that can be used with existing loss functions to enforce maximal area under the ROC curve (AUC) through prioritizing FPR reduction at high TPR. We solve the resulting constrained optimization problem using an Augmented Lagrangian method (ALM).Going beyond binary, we also propose two possible extensions of the proposed constraint for multi-class classification problems.We present experimental results for image-based binary and multi-class classification applications using an in-house medical imaging dataset, CIFAR10, and CIFAR100. Our results demonstrate that the proposed method improves the baselines in majority of the cases by attaining higher accuracy on critical classes while reducing the misclassification rate for the non-critical class samples.
Author Information
Sara Sangalli (ETH Zürich)
Ertunc Erdil (ETH Zurich)
Andeas Hötker (University Hospital Zurich)
Olivio Donati (University Hospital Zurich)
Ender Konukoglu (ETH Zurich)
More from the Same Authors
-
2023 Poster: Canonical normalizing flows for manifold learning »
Kyriakos Flouris · Ender Konukoglu -
2023 Poster: Expert load matters: operating networks at high accuracy and low manual effort »
Sara Sangalli · Ertunc Erdil · Ender Konukoglu -
2020 Workshop: Medical Imaging Meets NeurIPS »
Jonas Teuwen · Qi Dou · Ben Glocker · Ipek Oguz · Aasa Feragen · Hervé Lombaert · Ender Konukoglu · Marleen de Bruijne -
2020 Poster: Contrastive learning of global and local features for medical image segmentation with limited annotations »
Krishna Chaitanya · Ertunc Erdil · Neerav Karani · Ender Konukoglu -
2020 Oral: Contrastive learning of global and local features for medical image segmentation with limited annotations »
Krishna Chaitanya · Ertunc Erdil · Neerav Karani · Ender Konukoglu -
2019 Workshop: Medical Imaging meets NeurIPS »
Hervé Lombaert · Ben Glocker · Ender Konukoglu · Marleen de Bruijne · Aasa Feragen · Ipek Oguz · Jonas Teuwen -
2019 : Opening Remarks »
Hervé Lombaert · Ben Glocker · Ender Konukoglu · Marleen de Bruijne · Aasa Feragen · Ipek Oguz · Jonas Teuwen -
2018 : Closing remarks »
Ender Konukoglu · Ben Glocker · Hervé Lombaert · Marleen de Bruijne -
2018 : Welcome »
Ender Konukoglu · Ben Glocker · Hervé Lombaert · Marleen de Bruijne -
2018 Workshop: Medical Imaging meets NIPS »
Ender Konukoglu · Ben Glocker · Hervé Lombaert · Marleen de Bruijne -
2017 : Closing »
Ben Glocker · Ender Konukoglu · Hervé Lombaert · Kanwal Bhatia -
2017 : Poster session - Afternoon »
Yongchan Kwon · Young-geun Kim · Ender Konukoglu · Peter Li · John Guibas · Tejpal Virdi · Kuldeep Kumar · Morteza Mardani · Jelmer Wolterink · Enhao Gong · Natalia Antropova · Johannes Stelzer · Rene Bidart · Wei-Hung Weng · Martin Rajchl · Marc Górriz · Vineeta Singh · Christopher Sandino · Hiba Chougrad · Bob Hu · Isaac Godfried · Ke Xiao · Heliodoro Tejeda Lemus · Jordan Harrod · ILSANG WOO · Vincent Chen · Joseph Cheng · Vikash Gupta · Chuck-Hou Yee · Ben Glocker · Hervé Lombaert · Maximilian Ilse · Aneta Lisowska · Andrew Doyle · Milad Makkie -
2017 : Poster session - Morning »
Yongchan Kwon · Young-geun Kim · Ender Konukoglu · Peter Li · John Guibas · Tejpal Virdi · Kuldeep Kumar · Morteza Mardani · Jelmer Wolterink · Enhao Gong · Natalia Antropova · Johannes Stelzer · Rene Bidart · Wei-Hung Weng · Martin Rajchl · Marc Górriz · Vineeta Singh · Christopher Sandino · Hiba Chougrad · Bob Hu · Isaac Godfried · Ke Xiao · Heliodoro Tejeda Lemus · Jordan Harrod · ILSANG WOO · Vincent Chen · Joseph Cheng · Vikash Gupta · Chuck-Hou Yee · Ben Glocker · Hervé Lombaert · Maximilian Ilse · Aneta Lisowska · Andrew Doyle · Milad Makkie -
2017 : Opening »
Ben Glocker · Ender Konukoglu · Hervé Lombaert · Kanwal Bhatia -
2017 Workshop: Medical Imaging meets NIPS »
Ben Glocker · Ender Konukoglu · Hervé Lombaert · Kanwal Bhatia -
2017 : Poster Session »
Shunsuke Horii · Heejin Jeong · Tobias Schwedes · Qing He · Ben Calderhead · Ertunc Erdil · Jaan Altosaar · Patrick Muchmore · Rajiv Khanna · Ian Gemp · Pengfei Zhang · Yuan Zhou · Chris Cremer · Maria DeYoreo · Alexander Terenin · Brendan McVeigh · Rachit Singh · Yaodong Yang · Erik Bodin · Trefor Evans · Henry Chai · Shandian Zhe · Jeffrey Ling · Vincent ADAM · Lars Maaløe · Andrew Miller · Ari Pakman · Josip Djolonga · Hong Ge