Open AccessStellar Spectral Classification using Principal Component Analysis and Artificial Neural Networks
Author(s) -
Singh Harinder P.,
Gulati Ravi K.,
Gupta Ranjan
Publication year - 1998
Publication title -
monthly notices of the royal astronomical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.058
H-Index - 383
eISSN - 1365-2966
pISSN - 0035-8711
DOI - 10.1046/j.1365-8711.1998.01255.x
Subject(s) - principal component analysis , artificial neural network , curse of dimensionality , pattern recognition (psychology) , stellar classification , dimensionality reduction , artificial intelligence , physics , range (aeronautics) , dimension (graph theory) , spectral line , computer science , data mining , stars , astrophysics , mathematics , astronomy , materials science , pure mathematics , composite material
A fast and robust method of classifying a library of optical stellar spectra for O to M type stars is presented. The method employs, as tools: (1) principal component analysis (PCA) for reducing the dimensionality of the data and (2) multilayer back propagation network (MBPN) based artificial neural network (ANN) scheme to automate the process of classification. We are able to reduce the dimensionality of the original spectral data to very few components by using PCA and are able to successfully reconstruct the original spectra. A number of NN architectures are used to classify the library of test spectra. Performance of ANN with this reduced dimension shows that the library can be classified to accuracies similar to those achieved by Gulati et al. but with less computational load. Furthermore, the data compression is so efficient that the NN scheme successfully classifies to the desired accuracy for a wide range of architectures. The procedure will greatly improve our capabilities in handling and analysing large spectral data bases of the future.