t-SNE

t-SNE is a powerful dimensionality reduction technique used especially for visualizing complex datasets in two or three dimensions. Because it focuses on preserving local neighborhood structure, it can be very effective for revealing cluster-like patterns. However, the absolute geometry of its output is limited in meaning, and it is quite sensitive to hyperparameter choices. For that reason, t-SNE is most suitable for exploratory analysis and visual insight generation.