Causal Forecasting: How to disentangle causal effects, while controlling for ...
summary
Claims a deployable method that marries ML and econometrics to handle unobserved confounders by estimating causal elasticities, removing historical price effects, forecasting full‑price demand, then reapplying elasticities to simulate different discounts. Uses techniques like DoubleML, instrumental variables and panel interactive fixed effects alongside standard forecasting models. Presents a 3‑year developed solution tested on retail (pricing example where 40% discount gave ~3x uplift) and airline settings and shows the pipeline is operational for price optimization rather than mere forecasting.
AI-written summary · about
chapters (15)
- 0:00Causal forecasting introduction
- 1:06Speaker background
- 3:12Forecasting vs causal inference
- 6:04Unobserved confounders
- 8:15Pricing and confounders example
- 12:13Solving unobserved confounders
- 15:25Combining causal estimates with forecasting
- 18:58Causal forecasting workflow
- 22:29Practical implications
- 23:13Forecasting versus causality
- 24:55Summary
- 25:19Q&A
- 27:09Q&A: Local average treatment effect
- 29:00Q&A: Competitive pricing and counterfactuals
- 30:50Q&A: Open-source tools
related talks
43mBeyond Just Prediction: Causal Thinking in Machine Learning
06mCausal Machine Learning – What’s in it for Data Scientists?
39mCausal Effect Estimation in Practice: Lessons Learned from E-commerce & Banking
32mCausal inference and scenario generation within Just Eat Takeaway.com
11mCausal Inference For Socio-Economic & Engineering Systems
30mClimbing the causal ladder for fun and profit
16mUsing AI to solve complex business problems: optimizing...
31mFrom stockouts to happy customers: Proven solutions for time series forecasting in retail