Back to the program

Contributed Papers

Dynamics & Dependencies: Time Series and Copulas

Tue, Aug 4, 2:00 PM - 3:50 PM Room CC-255 Thomas M. Menino Convention & Exhibition Center

Jung Wun LeeChair

Section on Statistical Computing

7 Presentations

A Forward–Backward Framework for Imputation and Forecasting in Partially Observed Time Series

2:05 PM - 2:20 PM

Missing data imputation Time series Missing mechanism Electronic Health Records (EHRs)

Change point detection in variance

2:20 PM - 2:35 PM

Kyungduk Ko (Boise State University)

Change point Critical value Monte Carlo Size of test

COMPARING DIRECTIONAL MODELS FOR HIGH-DIMENSIONAL COMPOSITIONAL DATA

2:35 PM - 2:50 PM

Jose Guardiola (Texas A&M University-Corpus Christi)

Co-authors: Jacob Harris

Directional data Hyperspherical data Spherical Dirichlet distribution von Mises–Fisher Positive orthant movMF

Interval-Constrained Brownian Paths: Exact Interpolation and Extrapolation

2:50 PM - 3:05 PM

Radu Herbei (The Ohio State University)

Co-authors: Kumar Somnath (Foursquare Labs Inc.)

Brownian bridge Brownain meander exact simulation

Likelihood Inference in Gaussian Copula Models for Count Time Series via Minimax Exponential Tilting

3:05 PM - 3:20 PM

Quynh Nguyen (University of Texas at San Antonio)

Co-authors: Victor De Oliveira (University of Texas At San Antonio)

Durbin-Levinson algorithm GHK simulator High-dimensional integration Importance sampling Innovations algorithm

Model selection for a smooth empirical beta copula

3:20 PM - 3:35 PM

Xiaoling Dou (International Christian University)

Copula Cross-validation Empirical beta copula Model selection Smoothing

Pypomp: GPU-accelerated inference for partially observed Markov process models

3:35 PM - 3:50 PM

Aaron Abkemeier

Co-authors: Edward Ionides (University of Michigan), Jesse Wheeler (University of Michigan), Kevin En Tan, Jonathan Terhorst (University of Michigan), Aaron King (University of Michigan), Kunyang He (University of Michigan), Bo Yang (Columbia University), Jun Chen (Duke University)

partially observed Markov processes state space models sequential Monte Carlo statistical computing JAX maximum likelihood estimation

↑