Writing your own loss function for system identification

I'm working on a problem of system identification. The documentation for SI in 'Loss Function and Model Quality Metrics' illustrates a general form of the loss function used to estimate the parameters of the system. I was curious to know if we can write our own loss function which can be minimised by the solver?

Answer 1

[Ashwini Venkatappa]

The function to specify the original loss\evaluation function is not provided.