The causal problem of overexposure to repetitive ads: instrumental variables and pymc3
summary
Analyzes why targeted online advertising can lead to users being overexposed to repetitive ads by framing it as a causal inference problem. Uses a case study to show the causality issue and demonstrates applying instrumental variables together with Bayesian modeling in PyMC3 to address it.
AI-written summary · about
chapters (10)
- 0:00Greenhouse goals
- 1:49Ad metrics and caps
- 3:56Confounding in exposure
- 5:21Instrumental variables
- 7:56Frequency caps in tests
- 9:41Frequency cap distributions
- 11:25Causal effects with PyMC3
- 13:13Modeling variance
- 15:04Higher-order impression effects
- 16:09IV recap
related talks
27mUsing Causal thinking to make Media Mix Modeling
46mFrom correlation to causation: measuring ad effectiveness at scale
39mCausal Effect Estimation in Practice: Lessons Learned from E-commerce & Banking
30mClimbing the causal ladder for fun and profit
29mMedia Mix Modelling
01h 26mProbabilistic Programming and Bayesian Computing with PyMC
30mIs your marketing effective? Let Bayes decide!
22mCommunicating Algorithmic Process in Online Behavioral Advertising