|
Activity Number:
|
82
- Computer Experiments, Statistical Engineering, and Applications in Physical Sciences
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2018 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
|
Abstract #329544
|
|
|
Title:
|
Modeling of Sediment Mixing Using Dirichlet Process Mixtures
|
|
Author(s):
|
John Tipton* and Glenn Sharman and Sam Johnstone
|
|
Companies:
|
University Of Arkansas and University Of Arkansas and United States Geological Survey
|
|
Keywords:
|
Sediment mixing;
Dirichlet Process;
Bayesian Nonparametrics;
Mixture model
|
|
Abstract:
|
Deposits of sediment on the Earth's surface provide a record of global change through geologic time. Sediment is eroded from source locations (parents), transported, and ultimately deposited in sink locations (daughters). To the extent that parent sources produce sediment with distinguishable geochronologic ages, the distribution of ages of sediment particles observed in sinks allows for investigation of the mixing proportions of each of the parents that contributed sediment. To model the proportion of the daughter age distribution that comes from each of the parent distributions, we use a Bayesian nonparametric finite mixture of infinite Dirichlet processes. This model allows for estimation of the mixing proportions with associated uncertainty while making minimal assumptions. We also present an extension to the model whereby we reconstruct unobserved parent distributions using multiple daughter distributions using an infinite mixture of Dirichlet processes model, accounting for uncertainty in both the number of parent distributions and the mixing proportions.
|
Authors who are presenting talks have a * after their name.
