Addressing the Real-world Class Imbalance Problem in Dermatology

Wei-Hung Weng, Jonathan Deaton, Vivek Natarajan, Gamaleldin F. Elsayed, Yuan Liu
Proceedings of the Machine Learning for Health NeurIPS Workshop, PMLR 136:415-429, 2020.

Abstract

Class imbalance is a common problem in medical diagnosis, causing a standard classifier to be biased towards the common classes and perform poorly on the rare classes. This is especially true for dermatology, a specialty with thousands of skin conditions but many of which have low prevalence in the real world. Motivated by recent advances, we explore few-shot learning methods as well as conventional class imbalance techniques for the skin condition recognition problem and propose an evaluation setup to fairly assess the real-world utility of such approaches. We find the performance of few-show learning methods does not reach that of conventional class imbalance techniques, but combining the two approaches using a novel ensemble improves model performance, especially for rare classes. We conclude that ensembling can be useful to address the class imbalance problem, yet progress can further be accelerated by real-world evaluation setups for benchmarking new methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v136-weng20a, title = {Addressing the Real-world Class Imbalance Problem in Dermatology}, author = {Weng, Wei-Hung and Deaton, Jonathan and Natarajan, Vivek and Elsayed, Gamaleldin F. and Liu, Yuan}, booktitle = {Proceedings of the Machine Learning for Health NeurIPS Workshop}, pages = {415--429}, year = {2020}, editor = {Alsentzer, Emily and McDermott, Matthew B. A. and Falck, Fabian and Sarkar, Suproteem K. and Roy, Subhrajit and Hyland, Stephanie L.}, volume = {136}, series = {Proceedings of Machine Learning Research}, month = {11 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v136/weng20a/weng20a.pdf}, url = {https://proceedings.mlr.press/v136/weng20a.html}, abstract = {Class imbalance is a common problem in medical diagnosis, causing a standard classifier to be biased towards the common classes and perform poorly on the rare classes. This is especially true for dermatology, a specialty with thousands of skin conditions but many of which have low prevalence in the real world. Motivated by recent advances, we explore few-shot learning methods as well as conventional class imbalance techniques for the skin condition recognition problem and propose an evaluation setup to fairly assess the real-world utility of such approaches. We find the performance of few-show learning methods does not reach that of conventional class imbalance techniques, but combining the two approaches using a novel ensemble improves model performance, especially for rare classes. We conclude that ensembling can be useful to address the class imbalance problem, yet progress can further be accelerated by real-world evaluation setups for benchmarking new methods. } }
Endnote
%0 Conference Paper %T Addressing the Real-world Class Imbalance Problem in Dermatology %A Wei-Hung Weng %A Jonathan Deaton %A Vivek Natarajan %A Gamaleldin F. Elsayed %A Yuan Liu %B Proceedings of the Machine Learning for Health NeurIPS Workshop %C Proceedings of Machine Learning Research %D 2020 %E Emily Alsentzer %E Matthew B. A. McDermott %E Fabian Falck %E Suproteem K. Sarkar %E Subhrajit Roy %E Stephanie L. Hyland %F pmlr-v136-weng20a %I PMLR %P 415--429 %U https://proceedings.mlr.press/v136/weng20a.html %V 136 %X Class imbalance is a common problem in medical diagnosis, causing a standard classifier to be biased towards the common classes and perform poorly on the rare classes. This is especially true for dermatology, a specialty with thousands of skin conditions but many of which have low prevalence in the real world. Motivated by recent advances, we explore few-shot learning methods as well as conventional class imbalance techniques for the skin condition recognition problem and propose an evaluation setup to fairly assess the real-world utility of such approaches. We find the performance of few-show learning methods does not reach that of conventional class imbalance techniques, but combining the two approaches using a novel ensemble improves model performance, especially for rare classes. We conclude that ensembling can be useful to address the class imbalance problem, yet progress can further be accelerated by real-world evaluation setups for benchmarking new methods.
APA
Weng, W., Deaton, J., Natarajan, V., Elsayed, G.F. & Liu, Y.. (2020). Addressing the Real-world Class Imbalance Problem in Dermatology. Proceedings of the Machine Learning for Health NeurIPS Workshop, in Proceedings of Machine Learning Research 136:415-429 Available from https://proceedings.mlr.press/v136/weng20a.html.

Related Material