Predicting casing wear, riser wear, and friction factors in drilling operations may be achieved with data-driven models that use discrete inversion techniques to updated casing wear models, riser wear models, and/or friction factor models. For example, a method may applying a linear inversion technique or a nonlinear inversion technique to one or more parameters of at least one of a casing wear model, a riser wear model, or a friction factor model using historical data from a previously drilled well as input data to produce at least one of an updated casing wear model, an updated riser wear model, or an updated friction factor model, respectively; and implementing the at least one of the updated casing wear model, the updated riser wear model, or the updated friction factor model when designing and/or performing a drilling operation.