Joint Models for Handling Non-Ignorable Missing Data using Bayesian Additive Regression Trees: Application to Leaf Photosynthetic Traits Data

Yong Chen Goh, Wuu Kuang Soh, Andrew C. Parnell, Keefe Murphy · 2024

Read Joint Models for Handling Non-Ignorable Missing Data using Bayesian Additive Regression Trees: Application to Leaf Photosynthetic Traits Data by Yong Chen Goh, Wuu Kuang Soh, Andrew C. Parnell, Keefe Murphy (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-14946.txt.

Description

Missing data can bias a study when the gaps themselves carry meaning. Yong Chen Goh and co-authors build two joint models for that hard case. Both use a flexible tree-based method to model the outcomes, here leaf traits tied to photosynthesis. The first adds a standard model for the chance that a value is missing. The second models that chance with trees too, for cases where the link is bent or complex.

Simulations in single and multiple outcome settings tested both models against older methods. The pair recovered the true patterns better across many gap mechanisms. The team also applied them to real leaf trait data with climate measures. The work gives ecologists a sturdier way to learn from patchy trait records.

Catalog record

CREATORYong Chen Goh, Wuu Kuang Soh, Andrew C. Parnell, Keefe Murphy
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE79 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-14946

Browse the full library · About the Tower of Babel