UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction (2018-02-09T00:00:00.000000Z)

TL;DR

The UMAP algorithm is competitive with t-SNE for visualization quality, and arguably preserves more of the global structure with superior run time performance.

Abstract

UMAP (Uniform Manifold Approximation and Projection) is a novel manifold learning technique for dimension reduction. UMAP is constructed from a theoretical framework based in Riemannian geometry and algebraic topology. The result is a practical scalable algorithm that applies to real world data. The UMAP algorithm is competitive with t-SNE for visualization quality, and arguably preserves more of the global structure with superior run time performance. Furthermore, UMAP has no computational restrictions on embedding dimension, making it viable as a general purpose dimension reduction technique for machine learning.

Authors

Leland McInnes

1 papers

John Healy

1 papers

TL;DR

Abstract

Authors

References65 items

UMAP does not preserve global structure any better than t-SNE when using the same initialization

Visualizing structure and transitions in high-dimensional biological data

Deep learning multidimensional projections

Comparison Between UMAP and t-SNE for Multiplex-Immunofluorescence Derived Single-Cell Data from Tissue Sections

Dimensionality reduction for visualizing single-cell data using UMAP

BloodSpot: a database of healthy and malignant haematopoiesis updated with purified and single cell mRNA sequencing profiles

The art of using t-SNE for single-cell transcriptomics

Manifold learning of four-dimensional scanning transmission electron microscopy

(Self-Attentive) Autoencoder-based Universal Language Representation for Machine Translation

Deep convolutional Gaussian processes

Data-Driven Design: Exploring new Structural Forms using Machine Learning and Graphic Statics

Revealing multi-scale population structure in large cohorts

Human Bone Marrow Assessment by Single Cell RNA Sequencing, Mass Cytometry and Flow Cytometry

Fast Batch Alignment of Single Cell Transcriptomes Unifies Multiple Mouse Cell Atlases into an Integrated Landscape

Comprehensive analysis of retinal development at single cell resolution identifies NFI factors as essential for mitotic exit and specification of late-born cells

Gaussian mixture models with Wasserstein distance

Evaluation of UMAP as an alternative to t-SNE for single-cell data

Fast Interpolation-based t-SNE for Improved Visualization of Single-Cell RNA-Seq Data

Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms

Clustering with t-SNE, provably

A Molecular Census of Arcuate Hypothalamus and Median Eminence Cell Types

Visualizing Large-scale and High-dimensional Data

Numba: a LLVM-based Python JIT compiler

Distributed Representations of Words and Phrases and their Compositionality

Type 1 and 2 mixtures of Kullback-Leibler divergences as cost functions in dimensionality reduction based on similarity preservation

A Survey on Metric Learning for Feature Vectors and Structured Data

OMIP‐018: Chemokine receptor expression on human T helper cells

FlowRepository: A resource of annotated flow cytometry datasets associated with peer‐reviewed publications

Survey Article: An elementary illustrated introduction to simplicial sets

Efficient k-nearest neighbor graph construction for generic similarity measures

Classifying Clustering Schemes

Information Retrieval Perspective to Nonlinear Dimensionality Reduction for Data Visualization

Graph Laplacians and their Convergence on Random Neighborhood Graphs

Laplacian Eigenmaps for Dimensionality Reduction and Data Representation

Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering

A global geometric framework for nonlinear dimensionality reduction.

Mapping a Manifold of Perceptual Observations

Simplicial objects in algebraic topology

FUZZY SET THEORY AND TOPOS THEORY

A Nonlinear Mapping for Data Structure Analysis

Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis

Fit-sne

What do numbers look like?

Efficient Algorithms for t-distributed Stochastic Neighborhood Embedding

Category theory in context

Multicore-tsne. https://github.com/DmitryUlyanov/ Multicore-TSNE, 2016

Accelerating t-SNE using tree-based algorithms

Independent consultant

UCI machine learning repository

Shift-invariant similarities circumvent distance concentration in stochastic neighbor embedding and variants

A LEISURELY INTRODUCTION TO SIMPLICIAL SETS

Simplicial Homotopy Theory

METRIC REALIZATION OF FUZZY SIMPLICIAL SETS

Visualizing Data using t-SNE

Distance Metric Learning: A Comprehensive Survey

Di usion maps. Applied and computational harmonic analysis

Diffusion maps

Coifman and Stéphane Lafon . Di  usion maps

Pen-based recognition of handwrien digits data set. university of california, irvine

Columbia Object Image Library (COIL100)

object image library (coil-100. Technical report

Categories for the Working Mathematician

Analysis of a complex of statistical variables into principal components.

e MNIST database of handwri en digits

A Global Geometric Framework for Nonlinear Dimensionality Reduction

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names

Coifman and Stéphane Lafon . Di usion maps

Pen-based recognition of handwrien digits data set. university of california, irvine