It should be possible to define the shape of the crisp probability function that has some +- derivation value, We currently have a strict triangular function, but feedback from practice shows that some Gaussian curve would be helpful. The reason is that strong outliers of the probability are typically less common than small outliers.
Optically, this would mean a dropdown box for crisp probabilities in FuzzTress that offers a list of shape types. This type choice should be then communicating to the backend / analysis server.
It should be possible to define the shape of the crisp probability function that has some +- derivation value, We currently have a strict triangular function, but feedback from practice shows that some Gaussian curve would be helpful. The reason is that strong outliers of the probability are typically less common than small outliers.
Optically, this would mean a dropdown box for crisp probabilities in FuzzTress that offers a list of shape types. This type choice should be then communicating to the backend / analysis server.