A Step-by-Step Guide in detecting causal relationships using Bayesian Structure Learning in Python.

The use of machine learning techniques has become a standard toolkit to obtain useful insights and make predictions in many areas such as disease prediction, recommendation systems, natural language processing. Although good performances can be achieved, it is not straightforward to extract causal relationships with, for example, the target variable. In other words: which variables have a direct causal effect on the target variable? Such insights are important to determine the driving factors that reach the conclusion, and as such, strategic actions can be taken. A branch of machine learning is Bayesian probabilistic graphical models, also named Bayesian networks (BN), which can be used to determine such causal factors. Let's rehash some terminology before we jump into the technical details of causal models.