search

r-causal / causal-inference-in-R

Causal Inference in R: A book!
https://www.r-causal.org/
167 stars 42 forks source link

Challenges in Obtaining Valid Causal Effect Estimates with Machine Learning Algorithms #56

Open malcolmbarrett opened 1 year ago

malcolmbarrett commented 1 year ago

https://academic.oup.com/aje/advance-article/doi/10.1093/aje/kwab201/6322279 (code: https://github.com/amishler/nonparametricDoublyRobust)

related: https://github.com/yqzhong7/AIPW

tgerke commented 1 year ago

another good review of causal ML approaches just out in preprint

tgerke commented 1 year ago

The ML in mediation example we discussed: https://academic.oup.com/ectj/article/25/2/277/6517682 with R implementation in the medDML function of {causalweight} https://cran.r-project.org/web/packages/causalweight/index.html

malcolmbarrett commented 1 year ago

Cross fitting: https://arxiv.org/abs/2004.10337

tgerke commented 1 year ago

and an R bloggers example of cross-fitting! https://www.r-bloggers.com/2017/06/cross-fitting-double-machine-learning-estimator/

tgerke commented 1 year ago

holy moly even Jamie in on this game: https://economics.mit.edu/files/12538 and here's the R code: https://github.com/VC2015/DMLonGitHub/

See last paragraph of page 10 into page 11 that gives really good background with refs on cross-fitting, e.g.

A key device that we use to avoid strong entropy conditions is cross-fitting via sample splitting. Cross-fitting is a practical, efficient form of data splitting. Importantly, its use here is not simply as a device to make proofs elementary (which it does), but as a prac- tical method to allow us to overcome the overfitting/high-complexity phenomena that commonly arise in data analysis based on highly adaptive ML methods