Going to post interesting looking papers fresh (or slightly stale) off the arXiv and relevant journal RSS feeds here (hopefully on a fortnightly basis)
- Empirical bayes methods to quantify value of experimental results
- CATE-thresholding [essential step in policy learning: estimate DR scores, then subset to units exceeding a given threshold, say 0]
- Related to ^ ; QINI curves are central quantities in policy learning – constructing them for K-armed settings is useful
- Partial identification of effect heterogeneity using CDFs of Y^1 and Y^0
- Anytime valid confidence sequences for linear models (COI: these are my past and future colleagues)
- Updates to open textbook on ridge regression
- Optimisation tricks for Kernel Ridge Regression