Fitting the Continuum Component of A Composite SDSS Quasar Spectrum Using CMA-ES

Abstract: Fitting the continuum component of a quasar spectrum in UV/optical band is challenging due to contamination of numerous emission lines. Traditional fitting algorithms such as the least-square fitting and the Levenberg-Marquardt algorithm (LMA) are fast but are sensitive to initial values of fitting parameters. They cannot guarantee to find global optimum solutions when the object functions have multiple minima. In this work, we attempt to fit a typical quasar spectrum using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES). The spectrum is generated by composing a number of real quasar spectra from the Sloan Digital Sky Survey (SDSS) quasar catalog data release 3 (DR3) so it has a higher signal-to-noise ratio. The CMA-ES algorithm is an evolutionary algorithm that is designed to find the global rather than the local minima. The algorithm we implemented achieves an improved fitting result than the LMA and unlike the LMA, it is independent of initial parameter values. We are looking forward to implementing this algorithm to real quasar spectra in UV/optical band.