用非平衡最优传输度量提升降维与学习效果
Neighbor Embeddings Using Unbalanced Optimal Transport Metrics
- 引入非平衡最优传输的赫林格-柯朗尼科夫度量进行降维
- 在MedMNIST上分类准确率比传统方法提升81%
- 适合处理分布不均的数据，对医学图像分析有帮助
本文提出在降维与学习（监督和无监督）流程中使用非平衡最优传输（UOT）中的赫林格-柯朗尼科夫度量。在多个基准数据集（包括MedMNIST）上对比了UOT、常规最优传输（OT）及基于欧氏距离的方法。实验结果表明，经统计假设检验验证，UOT平均表现优于欧氏与OT方法。特别是在MedMNIST数据集上，分类任务中UOT优于OT达81%的时间；聚类任务中优于OT达83%，且在58%的情况下优于所有其他度量。
原文摘要 · Abstract (English)
This paper proposes the use of the Hellinger--Kantorovich metric from unbalanced optimal transport (UOT) in a dimensionality reduction and learning (supervised and unsupervised) pipeline. The performance of UOT is compared to that of regular OT and Euclidean-based dimensionality reduction methods on several benchmark datasets including MedMNIST. The experimental results demonstrate that, on average, UOT shows improvement over both Euclidean and OT-based methods as verified by statistical hypothesis tests. In particular, on the MedMNIST datasets, UOT outperforms OT in classification 81\% of the time. For clustering MedMNIST, UOT outperforms OT 83\% of the time and outperforms both other metrics 58\% of the time.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。