Hydraulic characteristics are among the most important features of the soil including retention lines of soil moisture and hydraulic conductivity. Hydraulic conductivity of the soil characterizes soil´ s ability to conduct water. It is not time consuming to measure hydraulic conductivity. On the contrary, humidity retention lines reflect the relations between humidity and humidity potential of soil. Their course is influenced by a number of factors such as grain and mineralogical composition, the humus content, bulk density and the structure of the soil. Measurement in the laboratory conditions is time consuming and financially costly, therefore, this work is focused on the alternative solutions to the detection of retention lines of the soil moisture.The aim of this diploma thesis was to create pedotransfer functions based on field and laboratory measurements which could be used to determine the humidity retention lines for the region of Myjava. In addition to pedotransfer functions, other methods were used as well. These methods have been processed in the programming language R. Specifically, it have been the method of multilinear regression and Gradient boosting machine method. The accuracy of the methods used was evaluated by correlation coefficient R. Methods have been compared and the method suitable to be used for the water management in practice have been evaluated.