Targeted-BEHRT: Deep learning for observational causal inference on longitudinal electronic health records.
Shishir RaoMohammad MamoueiGholamreza Salimi KhorshidiYikuan LiRema RamakrishnanAbdelaali HassaïneDexter CanoyKazem RahimiPublished in: CoRR (2022)
Keyphrases
- causal inference
- deep learning
- electronic health records
- clinical data
- additive noise
- unsupervised learning
- causal models
- medical records
- health care
- medical data
- abductive reasoning
- machine learning
- clinical trials
- nonmonotonic reasoning
- rare events
- discrete data
- observational data
- object recognition
- probabilistic model
- maximum likelihood
- knowledge discovery
- databases
- generative model
- decision trees
- co occurrence
- information extraction