|
Activity Number:
|
290
- Advanced Bayesian Topics (Part 3)
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2021 : 1:30 PM to 3:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #318741
|
|
|
Title:
|
WITHDRAWN An Interpretable and Identifiable Approach to Age Period Cohort Modeling
|
|
Author(s):
|
Sean Ryan and David S Matteson
|
|
Companies:
|
Cornell University and Cornell University
|
|
Keywords:
|
Functional Models;
Identifiable Inference;
Nonparametric models;
Dynamic Models;
Dual Time Dynamics
|
|
Abstract:
|
Age Period Cohort (APC) models are widely used to analyze incidences of disease across different age groups over time. These models can be very problematic in practice as linear trends in any of the parameters are unidentifiable, which the output of these models can be easily misinterpreted as a result. This has led to a significant focus In recent years on an identifiable parameterization of the model. Despite being identifiable, this approach is very challenging to interpret and fails to address typical inference questions. We propose a Bayesian dynamic nonparametric functional approach to APC modeling. This approach can be applied to a wide range of datasets with little to no tuning, provides a flexible yet interpretable representation of the data and allows researchers to robustly answer questions with their data/ Furthermore this approach allows for full uncertainty quantification. We show that our method compares favorably with current state of the art models and demonstrate its value on a number of real world datasets
|
Authors who are presenting talks have a * after their name.
