Parametric estimation of `SampleEntropy`.

[kahaaga]

Describe the feature you'd like to have

Aktaruzzaman & Sassi (2014) presents a parametric approach to computing SampleEntropy. It would be nice to have here.

Cite scientific papers related to the feature/algorithm

Aktaruzzaman, M., & Sassi, R. (2014). Parametric estimation of sample entropy in heart rate variability analysis. Biomedical Signal Processing and Control, 14, 141-147.

If possible, sketch out an implementation strategy

There are several alternatives:

• Making the estimation type a type parameter for SampleEntropy, let the SampleEntropy constructor take a keyword then dispatching on the estimator type.
• Alternatively, provide a complexity(measure::SampleEntropy, estimator::SampleEntropyEstimator, args...) method. I prefer this, because then we'll have the same syntax for entropy and complexity, which makes higher-level plug-in methods a lot easier to implement.

This is low priority. I haven't though the actual implementation through at all.

[Datseris]

This could be new complexity type all together, like SampleEntropyParametric so that it doesn't add weight to the interface by adding new API functions. A new API dispatch method signature is a new function. complexity(measure::SampleEntropy, estimator::SampleEntropyEstimator, args...) this is too specific to warrant a new dispatch documentation entry in the complexity function.