Performing symbolic regression (SR)

A mathematical model is a set of equations and parameters that return output from input. The mathematical model is always a compromise between precision and simplicity. In fact, it is useless to resort to sophisticated models when the values of the parameters that appear in them are known only approximately. The search for mathematical models can sometimes be very complicated – for example, when trying to symbolize a markedly nonlinear phenomenon. In these cases, the researcher can find valuable help from the process of extrapolating information and relationships present in the input data, in the form of a symbolic equation. In this case, it is a symbolic regression process, different from its classical counterpart in that it has the advantage of simultaneously modifying the underlying structure and the parameters of the mathematical model. A symbolic regression model can be represented by the following equation:

...