Forecasting Unemployment based on Fuzzy Time Series with Different Degree of Confidence

Abstract

Unemployment prediction has attracted much attention to many sectors as it can be a guide for decision making planning. In the last few years, many different forecasting techniques based on fuzzy time series (FTS) have been designed. However, most of the models used the discrete fuzzy set as a base for calculating the predicted values and the discrete fuzzy set cannot provide the forecasted range under different degree of confidence (DDoC). In this paper, FTS with trapezoidal fuzzy numbers and frequency density based partitioning approach is introduced for predicting unemployment. This model is an enhancement of the previous FTS models as it can produce the forecasted ranges under DDoC which can provide more information on the forecasted values.