Abstract
Probabilistic fuzzy set is used to model the non-probabilistic and probabilistic uncertainties simultaneously in the system. This study proposes a cumulative probability-based discretization and probabilistic fuzzy set based novel fuzzy time series forecasting method. It also proposes a novel discretization approach based on cumulative probability to tackle the probabilistic uncertainty in partitioning of datasets. Gaussian probability distribution function has been used to construct probabilistic fuzzy set. The advantage of the proposed work is that it addresses the uncertainties due to randomness and fuzziness simultaneously and also improves accuracy rate in time series forecasting. A proposed forecasting method is applied on two time series data set of enrollments of University of Alabama and Taiwan Exchange (TAIEX). A reduction of the amount of average forecasting error rate (AFER) and root mean square error (RMSE) shows the proposed method outperforms over other existing forecasting methods.
Recommended Citation
Gupta, Krishna Kumar and Saxena, Suneet
(2024).
A Novel Fuzzy Time Series Forecasting Method Based on Probabilistic Fuzzy Set and CPBD Approach,
Applications and Applied Mathematics: An International Journal (AAM), Vol. 19,
Iss.
3, Article 6.
Available at:
https://digitalcommons.pvamu.edu/aam/vol19/iss3/6