Title :
Modelling volatility with mixture density networks
Author :
Mostafa, Fahed ; Dillon, Tharam
Author_Institution :
DEBII, Curtin Univ., Bentley, WA
Abstract :
Volatility is an important variable in financial forecasting. Forecasting volatility requires a development of a suitable model for it. In this paper, we examine different time series models for volatility modelling. Specifically, we will study the use of recurrent mixture density networks, GARCH and EGARCH models to model volatility. In addition, we demonstrate the impact of different factors on the accuracy and completeness of each of these models.
Keywords :
financial management; recurrent neural nets; time series; exponential GARCH model; financial forecasting; recurrent mixture density network; time series; volatility modelling; Economic forecasting; Equations; Gaussian processes; Instruments; Neural networks; Portfolios; Predictive models; Pricing; Risk management; Testing;
Conference_Titel :
Granular Computing, 2008. GrC 2008. IEEE International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4244-2512-9
Electronic_ISBN :
978-1-4244-2513-6
DOI :
10.1109/GRC.2008.4664673
