1-Why Causal Inference ?
- Correlation does not equal to Causation:Just because two variables move together does not mean one causes the other;
- Intervention Planning : Helps determine the impact of treatments, policies , and business strategies;
- Counterfactual Reasoning : Answers questions like "What would happen if we had done X instead of Y?"
2-Key Concepts in Causal Inference
- Potential Outcomes Framework (Rubin Causal Model) ;
- Confounding & Selection Bias;
3-Methods for Causal Inference
- Randomized Controlled Trials (RCTs)[Gold Standard] ;
- Matching Methods (For Observational Data);
- Instrumental Variables (IV)[For Unmeasured Confounding];
- Difference-in-Differences(DiD) [For Policy Analysis];
- Regression Discontinuity Design (RDD) [For Cutoff-Based Interventions];
- Structural Causal Models (SCM)[Graph-Base
4-Challenges & Future Directions
- Hidden Confounders: unobserved variables can bias results ;
- Data Limitations : Causal inference methods often require large and well-structured datasets;
- Causal AI : Integrating causal reasoning into machine learning models;
- Automated Causal Discovery : Using algorithms to identify causal structures in big data;