Population stability index
The population stability index (PSI) is a measure of the difference between two probability distributions, often used to quantify shifts in a population or model inputs.
AI Model Evaluation
The Population Stability Index measures the degree of distribution shift between two datasets or populations.
- How it is used
- Uses distribution shift as an early warning that offline evaluation data may no longer represent production traffic.
- How deep it goes
- Explains thresholds, binning, alerts, and interpretation limits, but stops short of proving causality or model degradation.
- Examples
- Uses training-versus-production features, histogram bins, drift alerts, and dashboards combining PSI with CTR or retention.
- What you can do after
- Diagnose distribution drift with PSI, then validate its product impact through performance checks, segments, and domain knowledge.
Read how AI Model Evaluation explains population stability index →
Financial AI in Practice
The population stability index quantifies changes between distributions of a variable or model population across time or samples.
- How it is used
- Monitors applicant-population drift in credit scoring as customer demographics or economic conditions change.
- How deep it goes
- Identifies PSI as a drift-monitoring metric but does not specify its calculation, thresholds, or response procedure.
- Examples
- Flags applicant-population shifts associated with changing customer demographics or economic conditions.
- What you can do after
- Diagnose potential population drift by using PSI to flag changes in credit applicants.
Read how Financial AI in Practice explains population stability index →
Machine Learning System Design
The population stability index is a divergence measure that quantifies how a population's distribution differs from a reference distribution.
- How it is used
- Uses prediction drift as an early warning for possible future model-quality changes.
- How deep it goes
- Defines an operational threshold and compares PSI with Wasserstein distance, but defers final calibration to historical experimentation.
- Examples
- Illustrates monthly comparisons using growth-adjusted reference data, including a 15% increase between April 2021 and April 2022.
- What you can do after
- Operate prediction monitoring by testing PSI alerts alongside Wasserstein distance and validating thresholds against historical data.
Read how Machine Learning System Design explains population stability index →
Related concepts
- data driftmeasures · Data drift is a change in the statistical distribution or characteristics of data relative to a reference, such as data used to train a model.
- Evidently AImeasured by · Evidently AI is an open-source Python tool for evaluating, testing, monitoring, and reporting on machine-learning data, models, predictions, and behavior.
- ScorecardMonitoringused by · ScorecardMonitoring is a scorecard-monitoring component that evaluates changes in score distributions and related statistical indicators.
- Wasserstein distancealternative to · Wasserstein distance is a distance between probability distributions defined by the minimum cost of transporting probability mass between them under a specified ground distance.
- model calibrationuses · Model calibration is the assessment or adjustment of a model’s outputs, scores, or confidence so they correspond more reliably to observed outcomes or a reference standard.
Questions about population stability index
What is population stability index?
The population stability index (PSI) is a measure of the difference between two probability distributions, often used to quantify shifts in a population or model inputs.
Which Manning books explain population stability index?
AI Model Evaluation, Financial AI in Practice and Machine Learning System Design.
What is the difference between population stability index and Wasserstein distance?
The population stability index (PSI) is a measure of the difference between two probability distributions, often used to quantify shifts in a population or model inputs. Wasserstein distance is a distance between probability distributions defined by the minimum cost of transporting probability mass between them under a specified ground distance.
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.


