Document Type

Dissertation

Rights

Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence

Disciplines

Computer Sciences

Publication Details

A dissertation submitted in partial fulfilment of the requirements of Technological University Dublin for the degree of M.Sc. in Computing (Data Analytics), 2021.

Abstract

The year 1943 saw the introduction of the Morgan-Keenan (MK) classification scheme and this replaced the existing Harvard Classification scheme. Both stellar classification scheme are fundamentally grounded in the field of spectroscopy. The Harvard Classification scheme classified stars based on stellar surface temperature. The MK Classification scheme introduced the concept of a luminosity class that is intrinsically linked to the surface gravity of a star. Temperature and luminosity class values are estimated directly from the stellar spectrum.

Machine learning is a well-established technique in astronomy. Traditionally, a spectrum is treated as a one-dimensional sequence of data. Techniques such as artificial neural networks and principal component analysis are commonly used when classifying spectra. Recent research has seen the application of convolutional neural networks in this domain.

This research investigates the effectiveness of using convolutional neural networks with folded spectra. Robust experimental and statistical techniques were used to test this hypothesis. The result show that folded spectra and 2D convolutional neural networks obtained a higher average classification accuracy when compared to spectra processed with a 1D convolutional neural network. A ResNet V2 50 architecture was also included in this experiment, but the results show that it did not match the performance of shallower network architecture.

All data used in this research has been archived on github and is available by following this link https://github.com/D18124324/dissertation

DOI

https://doi.org/10.21427/setb-yd94


Share

COinS