Causal Discovery and Machine Learning – from Assumptions to Applications
In the previous chapter, we reviewed three classes of sources of causal knowledge and discussed their main advantages and disadvantages. In this chapter, we’ll focus on the last source of knowledge mentioned in the previous chapter – causal discovery.
We’ll start by reviewing the popular assumptions behind causal discovery. Next, we’ll present four broad families of methods for causal discovery and we’ll introduce gCastle – the main Python package that we’ll use in this chapter. We’ll follow with a comparison of selected methods and a practical guide on how to combine causal discovery algorithms with expert knowledge.
By the end of this chapter, you will know a broad range of causal discovery methods. You’ll be able to implement them using Python and gCastle and you’ll understand the mechanics and implications of combining selected...
