Seismic reflection data can be used to generate high-resolution in-depth images capable of facilitating, with high precision, the correct positioning of wells in hydrocarbon exploration and production. However, images produced by migrating seismic data are often contaminated by artifacts due to multiple internal reflections. Different schemes can be used to avoid problems caused by these noises and to attenuate them, such as the Marchenko multiple elimination scheme (MME). Various solutions based on the MME method have been proposed in the literature. Therefore, in this work, we explore the MME based on the least-squares schemes (LSMME), the MME as Neumann series approximation solution (NMME), and the MME based on beyond Neumann method (BNMME), and compare them in terms of effectiveness and efficiency in different numerical examples.