Add normalising transformation

[BradyPlanden]

Feature description

Normalising the parameter space to a common amplitude and scale can be viewed as a special-case for the ScaledTransformation instance. For more information, see https://en.wikipedia.org/wiki/Normalization_(statistics)

Motivation

This transformation can provide convergence and robustness improvements for the optimisation and sampling processes when hyper-parameters are not calibrated.

Possible implementation

Add an instance of the ScaledTransformation that scales the parameter space to an amplitude of 1 and range of [0,1].

Additional context

No response