This repository contains my notes on causal graphs. Specifically, defining causal graphs and how to apply causal graph concepts to improve the identification of cause-effect relationships. This method will likely assist with the design of machine learning algorithms with the prime objective of improving prediction accuracy with a secondary objective of improved explanations of the predictions. Ultimately, I envisage causal graphs being one of the main ways that data scientists will harness machine learning and artificial intelligence.
These concepts originate from Judea Pearl's work on the Theory of Causality, which provides a concise set of mathematical operations for causal learning.
See the References page for useful background information about causal graphs.