Contents
What is Tmle?
TMLE is a semi-parametric doubly robust method that improves the chances of correct model specification by allowing for flexible estimation using non-parametric machine-learning methods. It therefore requires weaker assumptions than its competitors.
What is targeted maximum likelihood estimation?
Targeted maximum likelihood estimation is an algorithm1 for constructing a substitution (or “plug-in”) estimator for a given parameter ψ, in a (often nonparametric or semiparametric) model . Here, by parameter, we generally mean a smooth2 function from the data generating distribution p to a real-valued vector ψ(p).
How does targeted maximum likelihood estimation ( tmle ) work?
Targeted Maximum Likelihood Estimation (TMLE) is a semiparametric estimation framework to estimate a statistical quantity of interest. TMLE allows the use of machine learning (ML) models which place minimal assumptions on the distribution of the data.
How is targeted maximum likelihood used in causal inference?
Targeted Learning is proposed by van der Laan & Rubin in 2006 [1] as an automated (as opposed to do-it-yourself) causal inference method. TMLE is used to analyze censored observational data from a non-controlled experiment in a way that allows effect estimation even in the presence of confounding factors.
Which is the maximum likelihood based substitution estimator?
This is what we call g-computation in causal inference, it is a maximum-likelihood-based substitution estimator, it relies on the estimation of the conditional expectation of the outcome given the exposure and covariance. This estimator is used to generate the potential outcome Y₁ and Y₀, corresponding to A=1 and A=0.
What can tmle be used for in statistics?
TMLE can be used to estimate various statistical estimands (odds ratio, risk ratio, mean outcome difference, etc.) even when causal assumptions are not met. TMLE is, as its name implies, simply a tool for estimation.