I think so that it may depend on the nature of the uncertainty.
Dear Karthick,
I suggest you to study the following research item. It can help you.
https://www.sciencedirect.com/science/article/abs/pii/0165011494900035
The most membership functions are triangular because they allow simple calculations that can be
understood by researchers in the diagnostic domain.
Gaussians can also be used, but they are preferred when
the rules are learned automatically.
So the use of membership function depends on your domain.
In general, it is not possible to say which membership function is best for solving your problem. In my opinion, you choose one of these functions as the optimal membership function by trial and error. However, Gaussian functions and their families are considered as soft calculations and are a better option for expressing uncertainty. But that does not mean it is a better choice to solve all the problems.
- Badges
- Science topic