27.08.2019 Views

book

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Plenary Talks<br />

CF6<br />

Thursday, September 5th<br />

17:00<br />

Stationary Count Time Series<br />

Robert Lund<br />

Clemson University<br />

This talk overviews the modeling of stationary count time series, detailing some history and recent<br />

breakthroughs. Classical work involving the discrete and integer autoregressive moving-average<br />

model classes is first reviewed. Drawbacks with these models are illuminated and used to motivate<br />

two more modern approaches: copulas and construction from stationary sequences of zeroes and<br />

ones. What emerges are very flexible model classes that are naturally parsimonious, can have<br />

negative autocorrelations and/or long-memory features, and can be statistically fitted by likelihood,<br />

composite likelihood, and/or moment methods. Various applications are pursued, including a<br />

bivariate hurricane count model with Poisson components that are negatively correlated.<br />

13

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!