denoising score matching does: learns to denoise, which equals learning the score

Propensity score matching (PSM) reduces bias in treatment effect estimates

Propensity score matching (PSM) is a technique introduced by Paul R. Rosenbaum and Donald Rubin in 1983. It aims to estimate the effect of an intervention by accounting for covariates that predict receiving the treatment, thus reducing bias caused by confounding variables.

Example

In a study comparing the effects of a new drug on heart disease, PSM can be used to match patients who received the drug with those who did not, based on observed covariates like age, gender, and smoking status. This helps ensure that the comparison between treated and untreated groups is fair and reduces bias in the estimated treatment effect.

Remember this

Understanding PSM is crucial for researchers to obtain accurate estimates of treatment effects in observational studies, where randomization is not possible.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

score matching does: learns the gradient of the log-density without normalizing

Propensity score matching reduces bias in treatment effect estimates

Boosting (machine learning)

Boosting reduces bias in ML models

Resampling (statistics)

Bootstrapping samples with replacement to estimate distributions

classifier-free guidance does: interpolates between conditional and unconditional generation

"Classifies samples as either conditioned or unconditioned, guiding generation towards desired outcomes."

Regression discontinuity design

RDD uses a sharp threshold for treatment assignment

Brier score

Brier score measures mean squared error of probability predictions

Swipe through 100 ML concepts daily

Open Pocket Polymath