Machine Learning

Causal Inference

Statistical methods for determining cause-and-effect relationships from data, going beyond correlation to understand whether X actually causes Y.

Why It Matters

Causal inference prevents costly mistakes — 'Users who buy premium also use feature X' does not mean forcing feature X will drive premium sales.

Example

Determining whether a new onboarding flow actually caused increased retention, or whether the improvement was just due to seasonal effects or a concurrent marketing campaign.

Think of it like...

Like a doctor determining whether a medication cured the patient versus the patient recovering naturally — correlation (took medicine and got better) does not prove causation.