Skip to content
ML Academy

Introduced on day 21

Forecasting workflow

The checklist of problem definition, information gathering, exploratory analysis, model fitting, forecasting and evaluation that frames a forecasting project.

In detail

The textbook lists five basic steps and merges "use" and "evaluate" into one; this course separates them so that producing forecasts from a fixed origin and judging them on rolling origins stay distinct. Each step has an artefact: a framing card, a data and driver inventory, an exploration note with dated breaks, candidate models with baselines, a forecast file with its origin, and a backtest report.