Computational econometrics
Much recent research within the discipline has been concerned with developing computationally intensive estimation methodology for econometric and statistical models. These involve the development of computer-intensive algorithms that exploit the availability of fast processors. Their development has been driven by the growth of high frequency business and microeconomic data sources that warrant detailed modelling to accurately capture the underlying generating processes. Such complex models cannot be readily estimated using traditional methods of estimation, and computationally intensive methods are required. These methods include the generalised method of moments, simulated maximum likelihood, the EM algorithm and particularly Bayesian Markov Chain Monte Carlo methods. These have been applied to a wide variety of models including: semiparametric regression and time series models for intra-day electricity load forecasting, spatial smoothing models for real estate prices, stochastic volatility models for foriegn exchange prices, stated preference models in marketing, covariance matrix estimation for stock returns and semiparametric choice models from health economics.
Organisations with research expertise
Staff Members with research expertise
Organisations with research interests
Staff Members with research interests