The standard penetration test (SPT) is commonly considered to determine the shear strength parameters of cohesive soils, the relative density of cohesionless soils, and the stiffness and deformation properties of both cohesive and cohesionless soils. This information is essential for designing foundations, retaining walls, embankments, and other structures. Moreover, the evaluation of the factor of safety against liquefaction and liquefaction potential index are important parameters to assess the liquefaction potential characteristics of soil. Determination of the liquefaction potential behavior of soil is a prime challenge for geotechnical engineers as it influences human lives and Civil Engineering structures. The liquefaction potential phenomenon of soil is assessed through the cyclic resistance ratio which is used in geotechnical engineering to describe the cyclic shear strength of the soil. The cyclic resistance ratio of soil is calculated using the standard penetration test. In the present investigation, an effort is made to predict the standard penetration number through machine learning (ML) techniques. The SPT data of 53 boreholes are considered in the analysis. From the field data, soil parameters namely, bulk density, cohesion and friction angle are utilized as input parameters. In the study, multi-gene genetic programming (MGGP) and artificial neural network (ANN) are used as ML techniques to predict the SPT number. The prediction capabilities of mentioned ML techniques are compared through statistical parameters namely, coefficient of regression (R2), Nash–Sutcliffe efficiency (NS), Index of Agreement (d) and Modified Index of Agreement (dModified). It is resulted that soil friction angle has maximum influence on the SPT number. It is perceived that MGGP is the optimum technique to predict the SPT number.