Re: [Tftb-help] how to measure the performance of tfd

[Eric Chassande-Mottin]

> If just from visual comparison, it is subjective.If from theory of
> bias,variance, or resolution, most performace comparison used in
> mathmatical form. I would like to know is there any other efficient way
> to compare the TFDs?

your question is ill-posed if it is formulated separately from a specific
problem. one cannot answer your question in *general*.  the objective of
the TFTB is primaly to provide analysis tools. their << performances >>
will depend on the specific problem at hand (probably estimation of a
random process in your case since you mention bias and variance). the
references of the tutorial includes several pointers which may help you in
your evaluation.

e.