Calibration of $b$-tagging algorithm of Variable Radius Track-Jets using Tag and Probe method with semileptonic $t\bar{t}$ events.

Flavor tagging, namely the identification of the parton flavor ($b/c/{others}$) initiating a jet, is one of the fundamental ingredients for many analyses in ATLAS. In particular $b$-tagging is a combination of Machine Learning algorithms aiming to identify jets originating from the bottom quark. The performances and the efficiency of these algorithms are evaluated on Monte Carlo simulations and then applied to data. For these reasons, we need data-to-MC calibrations for these algorithms. In general, to calibrate the $b$-tagging algorithm on $b$-jets we need a pure sample enough before the application of the tagging algorithms itself, for this reason, $t\bar{t}$ samples are suitable for this task. In this presentation, the calibration of $b$-tagging efficiency for Track-Jets with Variable Radius is presented. These are created by applying an anti-$k_t$ algorithm, with a variable radius, only to the tracks in the events. The calibration is performed using $t\bar{t}$ events with a single lepton in the final state using a Tag and Probe method. The results are combined with an orthogonal measurement of the $b$-tagging efficiency to provide the optimal coverage as a function of the $b$-jet transverse momentum.