Title: Classification of Pulsar Glitch Amplitudes using Extreme Deconvolution

Abstract: We carry out a classification of the glitch amplitudes of radio pulsars using Extreme Deconvolution technique based on the Gaussian Mixture Model, where the observed uncertainties in the glitch amplitudes $\Delta \nu/\nu$ are taken into account. Our dataset consists of 699 glitches from 238 pulsars. We then use information theory criteria such as AIC and BIC to determine the optimum number of glitch classes. We find that both AIC and BIC show that the pulsar glitch amplitudes can be optimally described using a bimodal distribution. The mean values of $\Delta \nu/\nu$ for the two components are equal to $4.79 \times 10^{-9}$ and $1.28 \times 10^{-6}$, respectively with standard deviation given by 1.01 and 0.55 dex. We also applied this method to classify the pulsar inter-glitch time intervals, and we find that AIC prefers two components, whereas BIC prefers a single component. The unified data set and analyses codes used in this work have been made publicly available.