Chapter 09 · Sustainability & AISustainability & AI

Predictive analytics

Meaning statusEstablishedSource recordDirect document linkedWhy these are different

Definition

The use of statistical and machine-learning techniques on historical data to forecast future outcomes or estimate unknown quantities.

References

Springer / StanfordHastie, Tibshirani & Friedman — The Elements of Statistical Learning

This source provides part of the technical or institutional basis for the definition.

IEADigitalisation and Energy (2017)

This source supports the explanation of how the term is applied, measured or governed in practice.

Overview

What it means

Predictive analytics spans regression-style statistical models to modern machine learning, evaluated by forecast accuracy on unseen data. Its sustainability value lies in anticipation: demand, yield, failure, risk — estimated early enough to act on.

How it is used

Applications include energy-demand and renewable-output forecasting, supply-chain disruption risk, credit and insurance climate-risk scoring, and predicting equipment maintenance needs to extend asset life.

Why it matters

Prediction converts sustainability management from reactive to anticipatory — but forecasts embed the past. Under changing climate and market conditions, validating predictive systems against drift is as important as their average accuracy.