Feature description

Compute the observed Fisher Information Matrix (FIM) for the given inputs. The FIM is computed using the square of the gradient, divided by the number of data points.

Motivation

Often it becomes necessary to know how sensitive the optimization problem is to changes in the parameters. Acquiring Hessian information can be challenging. FIM can be calculated from the gradient, and is already implemented for the Jax cost functions in PyBOP.

Possible implementation

No response

Additional context

No response