In this study, permeability variations of clay soil and removal rates of ions presented in the leachate are determined. Penalized linear regression (Lasso) and support vector regression methods are applied in order to model the relationship between the metal ions and the permeability. For this purpose, leachate is collected from Sanitary Landfill at Sile-Komurcuoda, Istanbul. Permeability of samples is determined via consolidated clays which are compacted via standard methods. The concentrations of Fe2+, Mn2+, Zn2+, Cu2+ and Pb2+ ions in influent and effluent of reactor are anayzed and the removal rate of these ions are calculated in order to detect removal ability of clay soil. An overall evolution of Fe2+, Mn2+, Zn2+, Cu2+ and Pb2+ parameters is that removal rate of clay soil which was compressed with standard methods and consolidated was found higher than that of clay soil compressed with standard compaction method. For prediction accuracy and interpretation purposes, two methods are considered for modelling the data. Both methods show good generalization capabilities.