使用对比学习的有监督随机近邻嵌入
Supervised Stochastic Neighbor Embedding Using Contrastive Learning
AI总结：
本研究将自监督对比学习扩展至全监督设置，提出一种有监督随机近邻嵌入方法，通过在低维空间中拉近同类样本、推开异类样本，有效利用标签信息以优化数据可视化。
摘要：
随机近邻嵌入（SNE）方法 t-SNE、UMAP 是两种最流行的用于数据可视化的降维方法。对比学习，尤其是自监督对比学习（SSCL），在从未标记数据中嵌入特征方面展现了巨大成功。SNE 和 SSCL 之间的概念联系已被利用。在本研究中，在保留数据集近邻信息的范围内，我们将自监督对比方法扩展到全监督设置，使我们能够有效利用标签信息。属于同一类别的样本簇在低维嵌入空间中被拉近，同时将不同类别的样本簇推开。
英文摘要：
Stochastic neighbor embedding (SNE) methods $t$-SNE, UMAP are two most popular dimensionality reduction methods for data visualization. Contrastive learning, especially self-supervised contrastive learning (SSCL), has showed great success in embedding features from unlabeled data. The conceptual connection between SNE and SSCL has been exploited. In this work, within the scope of preserving neighboring information of a dataset, we extend the self-supervised contrastive approach to the fully-supervised setting, allowing us to effectively leverage label information. Clusters of samples belonging to the same class are pulled together in low-dimensional embedding space, while simultaneously pushing apart clusters of samples from different classes.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
