|
Activity Number:
|
31
- Methodological Advancements in Biostatistics
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 30, 2017 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
ENAR
|
|
Abstract #322793
|
|
|
Title:
|
Tukey G-And-H Autoregressive Processes for Non-Gaussian Time Series
|
|
Author(s):
|
Yuan Yan* and Marc G. Genton
|
|
Companies:
|
KAUST and KAUST
|
|
Keywords:
|
Non-Gaussian ;
Heavy tails ;
Skewness ;
Time series ;
Tukey g-and-h transformation ;
Trans-Gaussian
|
|
Abstract:
|
We construct a flexible autoregressive model that can adapt to time series data with features of skewness and heavy tails. Our model is based on the Tukey g-and-h transformation. We develop efficient algorithms for parameter estimation, order selection and prediction for our non-Gaussian time series model. We illustrate the estimation and prediction performance of our new model by a simulation study. We demonstrate the advantage of our model on hourly wind speed data and compare it with prediction based on other transformations.
|
Authors who are presenting talks have a * after their name.
