2. Measurement inversion with a linear model The minimization problem is solved differently depending on whether the model output is linear or non-linear with respect to the parameters: in the 1 er case, there is an explicit solution to the problem; in the 2nd case, iterative non-linear programming (NLP) methods must be used to approximate the solution. 2.1 Linear ordinary least squares (OLS) estimator When the model is linear, its output is written as : ymo(