Abstract. Statistical techniques are quite frequently used in the data analysis of present day high-energy and astrophysics experiments. Generally these methods are employed because of the lack of knowledge about the mathematical dependence of the parameter of interest on the relevant measured variables. A tree-based classification method ``Random Forest'' has been employed to segregate Gamma-ray and hadron initiated atmospheric Cherenkov events detected by ground based gamma-ray telescopes. The basic concepts of Random Forest method and its application for TACTIC telescope will be discussed.