تقرير
EXACT: How to Train Your Accuracy
العنوان: | EXACT: How to Train Your Accuracy |
---|---|
المؤلفون: | Karpukhin, Ivan, Dereka, Stanislav, Kolesnikov, Sergey |
سنة النشر: | 2022 |
المجموعة: | Computer Science |
مصطلحات موضوعية: | Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition |
الوصف: | Classification tasks are usually evaluated in terms of accuracy. However, accuracy is discontinuous and cannot be directly optimized using gradient ascent. Popular methods minimize cross-entropy, hinge loss, or other surrogate losses, which can lead to suboptimal results. In this paper, we propose a new optimization framework by introducing stochasticity to a model's output and optimizing expected accuracy, i.e. accuracy of the stochastic model. Extensive experiments on linear models and deep image classification show that the proposed optimization method is a powerful alternative to widely used classification losses. Comment: Pattern Recognition Letters (2024) |
نوع الوثيقة: | Working Paper |
DOI: | 10.1016/j.patrec.2024.06.033 |
URL الوصول: | http://arxiv.org/abs/2205.09615 |
رقم الأكسشن: | edsarx.2205.09615 |
قاعدة البيانات: | arXiv |
DOI: | 10.1016/j.patrec.2024.06.033 |
---|