key concepts
Causal Graphs(DAGs)
Causal Markov Condition
Causal Markov Condition
A causal graph consist of :
- Nodes : Represent different factors in the system.
- Directed Edges : indicate causal influence (e.g., A to B means A causes B)
- No cycles (i.e., no feedback loops).
Causal Markov Condition
Causal Markov Condition
Causal Markov Condition
The Causal Markov Condition(CMC)is a fundamental assumption in causal inference that defines the relationship between causal structures and probablistic independence. It is the bridge between causality and statistics, allowing us to extract causal relationships from data using graphical models.
Confounders & Colliders
Causal Markov Condition
Confounders & Colliders
A confounder is a variable that causes both the independent variable(X) and the dependent variable (Y) , creating a spurious association between them.
A collider is a variable that is caused by two other variables (X and Y) , but does not cause them.