Fuzzy Time Series Methods for Forecasting

Abstract

The rapid growth of time-series data in domains such as energy, finance, and IoT has intensified the trade-off between model interpretability and forecasting performance. Fuzzy Time Series (FTS) methods offer transparent, rule-based forecasts but lack systematic comparison against modern neural and linear approaches across varied real-world settings.