Bayesian Spectral Deconvolution Based on Binominal Distribution

When analyzing spectral data, it is effective to use spectral deconvolution, in which data with multiple peaks are expressed as a sum of basis functions to estimate parameters. In Bayesian spectral deconvolution, selecting an appropriate probabilistic model based on the situation is important. In this paper, we propose a new method of Bayesian spectral deconvolution based on a binomial distribution. Our method can be used in situations where data are given as proportions, such as in the measurement of absorption spectra. In this study, we compare our method with a Bayesian spectral deconvolution method based on a Poisson distribution, which is generally used for count data, and show that our method is useful for certain situations. Moreover, we show the effectiveness of our method in various situations.