This paper proposes the use of teaching–learning-based optimization on the Kausel-Roesset stiffness matrix (TLBO-KRSM) for estimating the site-specific shear wave velocity profile ( \({V}_{s})\) in the active multichannel analysis of surface waves (MASW) test. Global search operations are carried out by teaching factor through the teaching–learning process in the TLBO algorithm to obtain converged solutions. The TLBO algorithm (free of control parameter) is expected to predict the \({V}_{s}\) profile more accurately than traditional optimization techniques (e.g., genetic algorithm (GA), differential evolution (DE), artificial bee colony optimization (ABCO), and particle swarm optimization (PSO)). The control parameters of GA, DE, ABCO, and PSO are not sufficiently calibrated, so there is a high possibility of misinterpreting the site-specific \({V}_{s}\) profile. It utilizes the wavelengths and phase velocities of the experimental fundamental mode dispersion curve to fix the search space of soil layer thicknesses and \({V}_{s}\) . The proposed TLBO-KRSM algorithm and the other optimization techniques are examined on real field data sets by performing the MASW tests at the IIT Gandhinagar campus. The MASW test results are validated with downhole seismic tests along with the four datasets from various literature studies. When comparing the misfit error and the site-specific \({V}_{s}\) profiles, the TLBO-KRSM algorithm has been found to be superior and requires less computational effort to locate the low misfit region to other optimization algorithms for estimating the \({V}_{s}\) profiles.
