The classical question of modeling flow around a cylinder and computing the resulting forces becomes substantially more complex when the background flow is unsteady and the cylinder is near a wall. This case corresponds directly to real-world applications for wave-driven forces on submarine cables or pipelines. Notably, the lift forces for a cylinder resting on a wall cannot be described by a Morison equation, as the force depends strongly on the history of the velocity in the wake of the cylinder. Certainly a Navier-Stokes solution for the flow is possible, but for the range of Reynolds numbers that is required for engineering applications, although straightforward, may require a non-trivial computational effort. Alternatively, the force is necessarily a function of the past time-series, in principle a straightforward task for a data-driven model. Indeed, flows around a cylinder make for an interesting fluid mechanics problem in a machine learning context. Using the experimental data of Aristodemo et al. [1] for comparison, in this work, we compare various machine learning approaches (e.g., Harris [2]) to existing semi-empirical Wake-type models [3], as well as a Navier-Stokes solution, to see the advantages and drawbacks of each [1] F. Aristodemo, G. R. Tomasicchio, and P. Veltri. New model to determine forces at on-bottom slender pipelines. Coastal Engineering, 58(3): 267-280, 2011. [2] J. C. Harris. Faster than real-time, phase-resolving, data-driven model of wave propagation and wave-structure interaction. Applied Ocean Research. 16 pp. (in revisions). [3] K. Kuznetsov, J. C. Harris, L. Asmar, N. Germain, F. Aristodemo. 2018. Modification of a Wake model for hydrodynamic forces on submarine cables with a rough seabed. In Proceedings of the 16th Journées de l'Hydrodynamique. (Marseille, France, November 2018), 12 pp.