Novel Data-analysis Methods

The IDDR has a unique approach to data analysis that combines two scientific traditions: causal reasoning and statistical learning

Inferring Causality

The IDDR embraces the tradition of causal reasoning in Health Sciences. From Bradford Hill’s principles of causality to more recent theorems on causal inference, our aim is not only to predict, but to understand what promotes remission of diabetes.

Improving Statistical Analysis

The IDDR embraces the use of statistical learning techniques - also known as machine learning - to analyze the data and obtain more accurate estimations of the target parameters.

Giving up the perfect and pursuing the better.

Instead of looking for the perfect treatment for diabetes, we aim to iteratively achieve better algorithms to treat the disease.  The goal is to maximize the probabilities that a patient can achieve remission of diabetes, given the available approved treatments and current scientific knowledge.