MLE: NMaximize vs FindDistributionParameters

[mthiel (Marco Thiel)]

See also:

http://mathematica.stackexchange.com/questions/6498/standard-errors-for-maximum-likelihood-estimates-in-finddistributionparameters

and

Confidence Interval and EstimatedDistribution

Is there an easy way of obtaining a confidence interval directly from EstimatedDistribution? In NonLinearModelFit there’s the MeanPredictionBands and the ParameterConfidenceIntervalTable. There are still, also, the MeanCI and VarianceCI functions available. But how to easily implement a CI analysis, on a distribution fitted by EstimatedDistribution (or FindDistributionParameters)?

Cheers,

M.