Application of Multivariate Analysis in Separation of Higgs Boson Signal at future e+e- colliders

Abstract

Even though the environment at future e+e- colliders is practically QCD background free, there is a large number of processes with high cross-sections and/or similar topology as the Higgs signal of interest. Maximization of the achievable precision of measurements in the Higgs sector and beyond calls for optimized event selection with respect to the statistical significance. This is where the Multivariate Analysis (MVA) is employed, separating the signal from numerous backgrounds on the basis of their kinematic and other properties. In this paper, we discuss the basics of MVA, its application and performance, in examples of several Higgs analyses done in our group using full simulation of the CLIC data.