“…Motivated by the successful performance of XGBoost (Chen & Guestrin 2016) in International Challenges on Machine Learning (Xu 2018), and animated by the many different kinds of results presented by Pashchenko et al (2017), Smirnov & Markov (2017), Bethapudi & Desai (2018), Abay et al (2018), van Roestel et al (2018), Saha et al (2018), Lam & Kipping (2018), Shu et al (2019), Liu et al (2019), Askar et al (2019), Calderon & Berlind (2019), Chong & Yang (2019), Jin et al (2019), Menou (2019), Plavin et al (2019), Wang et al (2019), Yi et al (2019), , Lin et al (2020), Hinkel et al (2020), Tamayo et al (2020) and Tsizh et al (2020), we decided to test how this kind of algorithm would perform specific tasks related to the treatment of time series in radio datasets of AGNs, such as light curves of quasars and BL Lacs. For this reason we selected two well- PKS 1921-293 (GHz) 4.8 1977-20121979-20118.0 1968-20121974-201114.5 1974-20121975-2011 The UMRAO datasets were acquired in frequencies of 4.8 GHz, 8.0 GHz and 14.5 GHz from radio sources PKS 1921-293 (OV 236) and PKS 2200+420 (BL Lacertae).…”