|
Activity Number:
|
408
- SPAAC Poster Competition
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Scientific and Public Affairs Advisory Committee
|
|
Abstract #329317
|
|
|
Title:
|
Bayesian Function Data Analysis for Weather Forecast
|
|
Author(s):
|
Duchwan Ryu* and Hao Shen
|
|
Companies:
|
Northern Illinois University and Northern Illinois University
|
|
Keywords:
|
Bayesian Functional Data Analysis; Seeming Unrelated Regression; Dynamically Weighted Particle Filter; Radial Basis Function Networks; Spatio-temporal model
|
|
Abstract:
|
We consider a multivariate version of sequential Monte Carlo method to forecast weather. At a given day, we model multivariate weather data with the radial basis functional networks and the seemingly unrelated regression over all locations. To capture the change of weather data by days, we use transition models. As an efficient method for the demand computation and massive data involved in the model, we use dynamically weighted particle filter. We apply the proposed methods for the weekly forecast of weather data and compare its performance to the given forecast data with respect to the error distribution over all locations.
|
Authors who are presenting talks have a * after their name.
