AI Article Synopsis
- Several methods exist for estimating causal effects, but most struggle with complex data types like images.
- The proposed Causal Multi-task Deep Ensemble (CMDE) framework effectively learns both shared and unique information from different groups within a study population.
- CMDE demonstrates superior performance over current state-of-the-art methods by efficiently managing high-dimensional data and providing reliable estimates of causal effects across various datasets and tasks.*
Article Abstract
A number of methods have been proposed for causal effect estimation, yet few have demonstrated efficacy in handling data with complex structures, such as images. To fill this gap, we propose Causal Multi-task Deep Ensemble (CMDE), a novel framework that learns both shared and group-specific information from the study population. We provide proofs demonstrating equivalency of CDME to a multi-task Gaussian process (GP) with a coregionalization kernel . Compared to multi-task GP, CMDE efficiently handles high-dimensional and multi-modal covariates and provides pointwise uncertainty estimates of causal effects. We evaluate our method across various types of datasets and tasks and find that CMDE outperforms state-of-the-art methods on a majority of these tasks.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC10759931 | PMC |
Publication Analysis
Top Keywords
Similar Publications
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!