# Inductive Global and Local Manifold Approximation and Projection

Source: [Inductive Global and Local Manifold Approximation and Projection](<https://www.sota2.com/research/paper/68fa739c1e209131cea5d968>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-03T05:00:04.568Z

Authors: Jungeum Kim, Xiao Wang

Published: 2024-06-12T11:22:27.000Z

## Abstract

Nonlinear dimensional reduction with the manifold assumption, often called manifold learning, has proven its usefulness in a wide range of high-dimensional data analysis. The significant impact of t-SNE and UMAP has catalyzed intense research interest, seeking further innovations toward visualizing not only the local but also the global structure information of the data. Moreover, there have been consistent efforts toward generalizable dimensional reduction that handles unseen data. In this paper, we first propose GLoMAP, a novel manifold learning method for dimensional reduction and high-dimensional data visualization. GLoMAP preserves locally and globally meaningful distance estimates and displays a progression from global to local formation during the course of optimization. Furthermore, we extend GLoMAP to its inductive version, iGLoMAP, which utilizes a deep neural network to map data to its lower-dimensional representation. This allows iGLoMAP to provide lower-dimensional embeddings for unseen points without needing to re-train the algorithm. iGLoMAP is also well-suited for mini-batch learning, enabling large-scale, accelerated gradient calculations. We have successfully applied both GLoMAP and iGLoMAP to the simulated and real-data settings, with competitive experiments against the state-of-the-art methods.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Dimensionality Reduction on Hierarchical Macro](<https://www.sota2.com/research/sota/dimensionality-reduction-on-hierarchical-macro.md>): Dimensionality Reduction; Hierarchical Macro; Trustworthiness: 99.7.

- [Dimensionality Reduction on Hierarchical Meso](<https://www.sota2.com/research/sota/dimensionality-reduction-on-hierarchical-meso.md>): Dimensionality Reduction; Hierarchical Meso; Trustworthiness: 0.997.

- [Dimensionality Reduction on Hierarchical Micro](<https://www.sota2.com/research/sota/dimensionality-reduction-on-hierarchical-micro.md>): Dimensionality Reduction; Hierarchical Micro; Trustworthiness: 0.997.

- [Dimensionality Reduction on Eggs](<https://www.sota2.com/research/sota/dimensionality-reduction-on-eggs.md>): Dimensionality Reduction; Eggs; Trustworthiness: 99.8.

- [Dimensionality Reduction on Scurve](<https://www.sota2.com/research/sota/dimensionality-reduction-on-scurve.md>): Dimensionality Reduction; Scurve; Trustworthiness: 0.998.

- [Dimensionality Reduction on Severed](<https://www.sota2.com/research/sota/dimensionality-reduction-on-severed.md>): Dimensionality Reduction; Severed; Trustworthiness: 0.999.

- [Dimensionality Reduction on SPHERES](<https://www.sota2.com/research/sota/dimensionality-reduction-on-spheres.md>): Dimensionality Reduction; SPHERES; Trustworthiness Score: 0.615.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2406.08097>)

[PDF](<https://arxiv.org/pdf/2406.08097.pdf>)
