Ngai Hang Chan

My research focuses on statistical inference for time series, that is, sequences of observations that exhibit temporal dependence. Most of my work has been in establishing rigorous statistical procedures to handle nonstationary time series data. Roughly speaking, a time series is considered stationary if it has reached a certain state of "statistical equilibrium." In many applications, data often exhibit nonstationary behavior. My main interest is in drawing useful information out of nonstationary time series data. Due to the "disequilibrium" nature of the data, the underlying probabilistic structure becomes extremely complicated. I am involved in developing a novel approach to handling critical phenomenon of this kind. This research has applications in various disciplines, including testing the random walk hypothesis of the stock market.

I also have an interest in the analysis of spatial-temporal dependent data. Spatial data arise when observations are made in various locations and those that are close to each other are thought to be more similar than those that are distant. The main job of a statistician is to extract the most crucial spatial information from the data. My work has been stimulated by the problem of trying to understand the spatial clustering phenomenon of sudden infant death syndrome.

Some Related Publications

Chan, N.H., Kadane, J.B., and Jiang, T. (1998). "On the time series analysis of the diurnal cycle of small scale turbulence," Environmentrics, 9, pp. 235-244.

Chan, N.H., and Palma, W. (1998). "State space modeling of long-memory time series," Ann. Stat., 26, pp. 719-740.

Chan, N.H., and Petris, G. (2000). "Recent developments in heteroskedastic financial series," in Chan, W.S., Li, W.K., and Tong, H. (eds.), Statistics and Finance: An Interface, pp. 169-184. Imperial College Press, London.

Basak, G., Chan, N.H., and Palma, W. (2001). "Approximation of a long-memory time series with an ARMA(1,1) model," Journal of Forecasting., in press.

Chan, N.H. (2001). Elements of Financial Time Series. Wiley, New York.