Fuzzy Systems as Universal Approximators
IEEE Transactions on Computers
A probabilistic and statistical view of fuzzy methods
Technometrics
Eliciting and analyzing expert judgment: a practical guide
Eliciting and analyzing expert judgment: a practical guide
Hi-index | 0.00 |
In this chapter, we look at the problem of developing a probability density function (PDF) with little or no numerical data and from that PDF generating a cumulative distribution function (CDF). We use the CDF to help us find a reliability estimate, or limits, on that estimate, for a determined value, for a given variable under study. We have used fuzzy logic and expert judgment to produce the required PDE The technique used here is similar to the fuzzy control techniques described in Chapter 8. One major difference is that instead of utilizing only a single value that would normally be produced by defuzzification of the "truncated" output membership functions, we use all the information that we have generated. This includes the entire truncated output membership function that represents the variance of the knowledge generated by the fuzzy rule system. In this chapter, we also use the probability controller technique, described in Chapter 8, to analyze the same problem.