Inverse propensity weighting
Inverse propensity weighting estimates causal effects or counterfactual outcomes by weighting observed outcomes inversely to the probability of the observed treatment or action given covariates or context.
AI Model Evaluation
Inverse propensity scoring estimates counterfactual outcomes by weighting observed outcomes inversely by the probabilities that generated their actions.
- How it is used
- Estimates candidate policy value offline from decisions and outcomes generated by a different policy.
- How deep it goes
- Works through probability ratios, aggregation, assumptions, and stability checks, with a multi-metric implementation.
- Examples
- Uses item recommendation probabilities, movie engagement metrics, and an agent’s tool-routing decisions.
- What you can do after
- Evaluate and compare candidate policies offline, check estimate reliability, and identify results for online confirmation.
Read how AI Model Evaluation explains inverse propensity weighting →
Causal Inference for Data Science
Inverse probability weighting estimates causal effects by weighting observations by the inverse probability of receiving their observed treatment.
- How it is used
- Calculates average treatment effects from previously computed propensity scores.
- How deep it goes
- Connects weighting to the adjustment formula and mentions implementations, while deferring the full derivation as optional.
- Examples
- Uses patient-level treatment, outcome, confounder, and propensity-score tables.
- What you can do after
- Apply propensity scores to estimate average treatment effects from observational patient data.
Read how Causal Inference for Data Science explains inverse propensity weighting →
Also covered in
What to know first
- propensity scoreA propensity score is the estimated conditional probability that a unit receives a particular treatment or action given observed covariates or context.
- varianceVariance is the expected squared deviation of a random variable from its mean and measures the dispersion of its values.
Related concepts
- counterfactual reasoningused by · Counterfactual reasoning examines hypothetical outcomes under conditions, actions, or policies different from those that actually occurred.
- off-policy evaluationuses · Off-policy evaluation estimates the performance of a target policy using interaction data collected under a different behavior policy.
- resultuses · A result is an outcome, observation, or value produced or observed by an action, event, process, or evaluation.
- probability theoryuses · Probability theory is the mathematical framework for quantifying and reasoning about uncertainty in events, variables, and outcomes.
- propensity scoreuses · A propensity score is the estimated conditional probability that a unit receives a particular treatment or action given observed covariates or context.
- propensity score methodused by · A propensity-score method is a causal-inference technique that estimates treatment effects by matching, weighting, stratifying, or otherwise adjusting units according to their observed treatment probabilities.
Questions about inverse propensity weighting
What is inverse propensity weighting?
Inverse propensity weighting estimates causal effects or counterfactual outcomes by weighting observed outcomes inversely to the probability of the observed treatment or action given covariates or context.
Which Manning books explain inverse propensity weighting?
AI Model Evaluation and Causal Inference for Data Science.
What should I know before learning inverse propensity weighting?
Manning books that teach inverse propensity weighting usually introduce propensity score and variance first.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

