Standard machine learning methods are powerful prediction tools, but they cannot be deployed for causal inference without putting additional structure on the estimation problem. This course provides a practical introduction to causal machine learning. We discuss the difference between predictive and causal machine learning. We introduce methods that allow to control for high-dimensional confounders (double selection procedure, debiased/double machine learning). We estimate heterogeneous effects of policy and business interventions (causal forest). Furthermore, we consider optimal decision rules (reinforcement learning, bandit algorithms). We solve real-world economic and business problems in practical coding R session.