Here is a handy introduction to the basic principles of machine learning: Supervised vs Unsupervised Learning – What’s the Difference?. While reading it I paused to wonder whether there is any way to use machine learning for analyzing causal relationships. Machine learning automates the creation and evaluation of models but it doesn’t identify confounding variables, to run robustness checks, etc. I imagine that the predictions based on ML results tend to be rather conservative since they extrapolate the present to the future, rather than use a series of past events to identify root causes. Please feel free to correct me if there are already ML tools that have causal research functionality!
Lucas Jerusalimiec 