Reducing the Dimensionality of Data: Locally Linear Embedding of Sloan Galaxy Spectra

Abstract

We introduce Locally Linear Embedding (LLE) to the astronomical community as a new classification technique, using SDSS spectra as an example data set. LLE is a nonlinear dimensionality reduction technique which has been studied in the context of computer perception. We compare the performance of LLE to well-known spectral classification techniques, e.g. principal component analysis and line-ratio diagnostics. We find that LLE combines the strengths of both methods in a single, coherent technique, and leads to improved classification of emission-line spectra at a relatively small computational cost. We also present a data subsampling technique that preserves local information content, and proves effective for creating small, efficient training samples from a large, high-dimensional data sets. Software used in this LLE-based classification is made available.

51 pages, 12 figures, submitted to the Astronomical Journal. For associated code, see http://ssg.astro.washington.edu/software

References in corpus (2)

• The Seventh Data Release of the Sloan Digital Sky Survey
• Reliable Eigenspectra for New Generation Surveys

Cited by in corpus (28)

• Kepler Eclipsing Binary Stars. III. Classification of Kepler Eclipsing Binary Light Curves with Locally Linear Embedding
• The weirdest SDSS galaxies: results from an outlier detection algorithm
• Spectroscopic Needs for Imaging Dark Energy Experiments: Photometric Redshift Training and Calibration
• Automatic unsupervised classification of all SDSS/DR7 galaxy spectra
• Dimensionality Reduction of SDSS Spectra with Variational Autoencoders
• Exploring the Morphology of RAVE Stellar Spectra
• The GALAH survey: Chemical Tagging of Star Clusters and New Members in the Pleiades
• A review of unsupervised learning in astronomy
• ZFOURGE: Using Composite Spectral Energy Distributions to Characterize Galaxy Populations at 1<z<4
• Classification of Stellar Spectra with LLE
• A data-driven model for spectra: Finding double redshifts in the Sloan Digital Sky Survey
• Qualitative interpretation of galaxy spectra
• Chromospherically Active Stars in the RAVE Survey. I. The Catalogue
• Chromospherically Active Stars in the RAVE Survey. II. Young dwarfs in the Solar neighborhood
• Machine learning in APOGEE: Unsupervised spectral classification with $K$-means
• Do galaxies form a spectroscopic sequence?
• Locally linear embedding: dimension reduction of massive protostellar spectra
• Machine learning in APOGEE: Identification of stellar populations through chemical abundances
• On the fundamentality of the radial acceleration relation for late-type galaxy dynamics
• Single parameter galaxy classification: The Principal Curve through the multi-dimensional space of galaxy properties
• De-noising of galaxy optical spectra with autoencoders
• Mapping the Similarities of Spectra: Global and Locally-biased Approaches to SDSS Galaxy Data
• Galaxy Morphological Classification with Manifold Learning
• Galaxy Manifold: Characterizing and understanding galaxies with two parameters
• Data-Space Validation of High-Dimensional Models by Comparing Sample Quantiles
• Generating on-the-fly large samples of theoretical spectra through N-dimensional grid
• Empirically-Driven Multiwavelength K-corrections At Low Redshift
• A machine learning approach to assessing the presence of substructure in quasar host galaxies using the Hyper Suprime-Cam Subaru Strategic Program