skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: CLASSIFICATION OF STELLAR SPECTRA WITH LOCAL LINEAR EMBEDDING

Abstract

We investigate the use of dimensionality reduction techniques for the classification of stellar spectra selected from the Sloan Digital Sky Survey. Using local linear embedding (LLE), a technique that preserves the local (and possibly nonlinear) structure within high-dimensional data sets, we show that the majority of stellar spectra can be represented as a one-dimensional sequence within a three-dimensional space. The position along this sequence is highly correlated with spectral temperature. Deviations from this 'stellar locus' are indicative of spectra with strong emission lines (including misclassified galaxies) or broad absorption lines (e.g., carbon stars). Based on this analysis, we propose a hierarchical classification scheme using LLE that progressively identifies and classifies stellar spectra in a manner that requires no feature extraction and that can reproduce the classic MK classifications to an accuracy of one type.

Authors:
; ;  [1]; ;  [2]
  1. Astronomy Department, University of Washington, Box 351580, U.W. Seattle, WA 98195-1580 (United States)
  2. School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213 (United States)
Publication Date:
OSTI Identifier:
22034327
Resource Type:
Journal Article
Journal Name:
Astronomical Journal (New York, N.Y. Online)
Additional Journal Information:
Journal Volume: 142; Journal Issue: 6; Other Information: Country of input: International Atomic Energy Agency (IAEA); Journal ID: ISSN 1538-3881
Country of Publication:
United States
Language:
English
Subject:
79 ASTROPHYSICS, COSMOLOGY AND ASTRONOMY; ABSORPTION SPECTRA; ACCURACY; ASTRONOMY; ASTROPHYSICS; CARBON STARS; CLASSIFICATION; DATA ANALYSIS; EMISSION SPECTRA; GALAXIES; NONLINEAR PROBLEMS; ONE-DIMENSIONAL CALCULATIONS; THREE-DIMENSIONAL CALCULATIONS

Citation Formats

Daniel, Scott F, Connolly, Andrew, Vanderplas, Jake, Schneider, Jeff, and Liang, Xiong. CLASSIFICATION OF STELLAR SPECTRA WITH LOCAL LINEAR EMBEDDING. United States: N. p., 2011. Web. doi:10.1088/0004-6256/142/6/203.
Daniel, Scott F, Connolly, Andrew, Vanderplas, Jake, Schneider, Jeff, & Liang, Xiong. CLASSIFICATION OF STELLAR SPECTRA WITH LOCAL LINEAR EMBEDDING. United States. https://doi.org/10.1088/0004-6256/142/6/203
Daniel, Scott F, Connolly, Andrew, Vanderplas, Jake, Schneider, Jeff, and Liang, Xiong. 2011. "CLASSIFICATION OF STELLAR SPECTRA WITH LOCAL LINEAR EMBEDDING". United States. https://doi.org/10.1088/0004-6256/142/6/203.
@article{osti_22034327,
title = {CLASSIFICATION OF STELLAR SPECTRA WITH LOCAL LINEAR EMBEDDING},
author = {Daniel, Scott F and Connolly, Andrew and Vanderplas, Jake and Schneider, Jeff and Liang, Xiong},
abstractNote = {We investigate the use of dimensionality reduction techniques for the classification of stellar spectra selected from the Sloan Digital Sky Survey. Using local linear embedding (LLE), a technique that preserves the local (and possibly nonlinear) structure within high-dimensional data sets, we show that the majority of stellar spectra can be represented as a one-dimensional sequence within a three-dimensional space. The position along this sequence is highly correlated with spectral temperature. Deviations from this 'stellar locus' are indicative of spectra with strong emission lines (including misclassified galaxies) or broad absorption lines (e.g., carbon stars). Based on this analysis, we propose a hierarchical classification scheme using LLE that progressively identifies and classifies stellar spectra in a manner that requires no feature extraction and that can reproduce the classic MK classifications to an accuracy of one type.},
doi = {10.1088/0004-6256/142/6/203},
url = {https://www.osti.gov/biblio/22034327}, journal = {Astronomical Journal (New York, N.Y. Online)},
issn = {1538-3881},
number = 6,
volume = 142,
place = {United States},
year = {Thu Dec 15 00:00:00 EST 2011},
month = {Thu Dec 15 00:00:00 EST 2011}
}