Apply three membership-based principles to translate fuzzy forecast outputs into numeric values¶

Strategy · All strategies
Evidence · no claims cited

Description¶

Because model outputs are fuzzy sets, the article prescribes three interpretation principles for obtaining quantitative forecasts. If an output has only one maximum membership, select the midpoint of the corresponding interval; if there are two or more consecutive maximums, select the midpoint of the corresponding consecutive intervals; otherwise compute a weighted "mean" using standardized memberships and interval midpoints. "Using human subjective experience knowledge isone ofthe most important features of fuzzy forecasting methods."

Design Implications¶

Context¶

Requirements¶

• Output memberships from the fuzzy model, the defined intervals and their midpoints, and human subjective experience knowledge in interpretation.

Constraints¶

• The principles are presented specifically for interpreting the output of the fuzzy time series enrollment model.

Target Learners¶

• institutional researchers converting fuzzy forecasts into enrollment numbers

Target Learning Goals¶

• translating fuzzy model outputs into usable numeric enrollment predictions

Affordances¶

• Fuzzy Time Series First Order Time Invariant Model

Related Strategies¶

Examples¶

-

Key Sources¶

• Song, Qiang; Chissom, Brad S. (1991). Forecasting Enrollments with Fuzzy Time Series. https://eric.ed.gov/?id=ED340733