Real-world data (RWD) and real-world evidence (RWE) are playing an increasing role in health care decisions, and especially in allowing treatment effect estimation in real life settings. In comparative real world studies, confounding bias is generally considered to be the greatest threat to results validity. Common methods such as matching, propensity scores, and g-estimation are well-known tools to balance measured confounding factors and obtain valid treatment effect estimation.
In Targeted Learning framework, Targeted Maximum Likelihood Estimation (TMLE) is a well-established alternative method with desirable statistical properties. It is a semiparametric double‐robust method that improves the chances of correct model specification by allowing for flexible estimation using (nonparametric) machine‐learning methods. It therefore requires weaker assumptions than its competitors. 
The main objective of this internship will be to acquire knowledge about TMLE, to evaluate the method through simulations and to perform an application on an internal real world data study.
 Schuler, M. S., & Rose, S. (2017). Targeted maximum likelihood estimation for causal inference in observational studies. American journal of epidemiology, 185(1), 65-73.