Supervised Stochastic Neighbor Embedding Using Contrastive Learning
This work addresses the need for better visualization of labeled datasets in machine learning, but it appears incremental as it builds on existing connections between stochastic neighbor embedding and contrastive learning.
The paper tackles the problem of improving dimensionality reduction for data visualization by extending self-supervised contrastive learning to a fully-supervised setting, resulting in clusters of same-class samples being pulled together and different-class clusters pushed apart in the embedding space.
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.