Rubin–Neyman Causal framework for Probability Measures
A mathematical view of causal inference through lens of functionals.
Notes by
Disclaimer
This page contains archived lecture notes and research ideas in Statistics, Machine Learning, and Generative AI. All opinions are my own and do not represent any affiliated organizations.
Lastest Posts
Knockoffs: A framework of Control Variable Selection
How synthetic negative controls give finite-sample FDR control for variable selection.
Importance-Weighted Methods: From Sampling to Modern AI
How reweighting connects importance sampling, causal adjustment, clinical trial comparison, class-imbalance remedies, coreset selection, and deep generative ...
Bayesian Survival Analysis: Then and Now
How survival analysis evolved from censored-time likelihoods to generative AI survival models.
Optimal Transport and Schrödinger Bridge
Transforming probability distributions through geometry, entropy, and variational optimality
Bayesian Uncertainty from Martingale Posteriors
Defining Bayesian uncertainty without the prior and likelihood
Conformal Prediction & Inferential Tests
Distribution-free uncertainty quantification in the age of Deep Learning
Variational Bayes & Inference
The most important statistical-physics principle in the age of Generative AI