An Analysis Framework for Fuzzy Time Series Forecasting Models

Time series has been catching considerable attention due to its wide range of applications. Fuzzy logic concepts have been applied to the analysis of time series resulting in producing Fuzzy Time Series (FTS). The classical time series uses numbers whereas FTS uses fuzzy sets or linguistic values. FTS forecasting is effective when the inputs are linguistic characterized by imprecision in nature. Forecasting in the presence of multiple factors is very important and challenging at the same time. Many FTS forecasting models have been developed and presented in the literature. However, there are still some challenges and gaps that needs to be addressed. To identify these gaps, we developed an analysis framework to allow for a systematic evaluation of FTS forecasting models using a set of criteria. We analyzed prominent FTS forecasting models and identified a set of gaps yet to be addressed. The set of gaps is meant to serve as an eye-opener on issues to be addressed in future research.