“…Most frequently employed data pre-processing techniques to overcame the small size problem are liner and nonlinear Principal component analysis (Jalali, Mallipeddi and Lee, 2017;Feng, Zhou and Dong, 2019;Athanasopoulou, Papacharalampopoulos and Stavropoulos, 2020), Discriminant analysis (Abu Zohair, 2019;Li et al, 2019;Silitonga et al, 2021), Data augmentation (Han, Liu and Fan, 2018;Hagos and Kant, 2019;Fong et al, 2020), Virtual sample (Gong et al, 2017;MacAllister, Kohl and Winer, 2020;Zhu et al, 2020), Feature extraction (Kumar et al, 2018;Dai et al, 2020) and Auto-encoder (Feng, Zhou and Dong, 2019;Pei et al, 2021).…”