Face and Expression Recognition

RATS: Unsupervised manifold learning using low-distortion alignment of tangent spaces

This paper introduces a novel dimensionality reduction technique that generates low-dimensional embeddings while preserving the global structure within the data for a variety of biological and non-biological datasets.

Approximate Nearest Neighbor Graph Provides Fast and Efficient Embedding with Applications in Large-scale Biological Data

This paper introduces an improved non-linear dimension reduction algorithm using HNSW for nearest neighbor search and applies it to visualize large-scale biological data with competitive accuracy.

Training a neural network to learn other dimensionality reduction removes data size restrictions in bioinformatics and provides a new route to exploring data representations

The paper uses deep learning with t-SNE to analyze large datasets and overcome size limitations in bioinformatics.

Distance preserving dimension reduction withlocal-topology based scaling for improvedclassification of Biomedical data-sets.

This paper introduces TPDS, a dimension reduction method that preserves both distance and local topology, improving classification and handling outliers in biomedical datasets.

Using global t-SNE to preserve inter-cluster data structure

This paper introduces global t-SNE (g-SNE), a modified t-SNE method that preserves global inter-cluster structure during dimensionality reduction, and shows it outperforms standard t-SNE on synthetic and real datasets.