Cluster and then Embed: A Modular Approach for Visualization
By: Elizabeth Coda, Ery Arias-Castro, Gal Mishne
Potential Business Impact:
Shows data groups clearly, without messing up the big picture.
Dimensionality reduction methods such as t-SNE and UMAP are popular methods for visualizing data with a potential (latent) clustered structure. They are known to group data points at the same time as they embed them, resulting in visualizations with well-separated clusters that preserve local information well. However, t-SNE and UMAP also tend to distort the global geometry of the underlying data. We propose a more transparent, modular approach consisting of first clustering the data, then embedding each cluster, and finally aligning the clusters to obtain a global embedding. We demonstrate this approach on several synthetic and real-world datasets and show that it is competitive with existing methods, while being much more transparent.
Similar Papers
The Tree-SNE Tree Exists
Machine Learning (Stat)
Shows data patterns at different zoom levels.
Low-dimensional embeddings of high-dimensional data
Machine Learning (CS)
Makes complex data easier to understand and use.
DREAMS: Preserving both Local and Global Structure in Dimensionality Reduction
Machine Learning (CS)
Shows hidden patterns in complex data.