Predictive Analytics

Predictive analytics uses historical data with statistical methods and machine learning to estimate what is likely to happen. A model learns patterns from past records, such as how sales rise before a festival or which signs come before a customer leaves, and applies them to current data to produce forecasts or probabilities.

Common uses include demand forecasting, churn prediction, credit risk scoring and predictive maintenance. Predictions carry uncertainty, so good practice is to report a range and to test the model on data it has not seen.

For example, a distributor forecasts next month's demand for each product from three years of orders, so it can plan stock before shelves run empty.