Evolving Participatory Learning Fuzzy Modeling for Yield Curve Forecasting with Time-Varying Volatility

This paper suggests a dynamic approach for the term structure of interest rates forecasting using evolving participatory learning fuzzy modeling (ePL). The model includes a time-varying volatility structure in order to predict the yield curve factors. Thus, this framework both comprises an adaptive framework for term structure parameters behavior and deal with the uncertainty related to these factors by describing its variability patterns. Results based on US Treasury market data indicate that the ePL model outperformed a common approach in the literature, based on autoregressive processes, for shortand long-term horizons considering the fitness accuracy.