What it means
Time-series forecasting estimates future values from ordered observations, trends, seasonality, and related signals.
Why it matters
Structural breaks, delayed data, leakage from the future, and changing seasonality can invalidate a model that once performed well.
Practical focus
- Define the approved purpose, accountable owner, users, operating environment, and acceptable risk for time-series forecasting before implementation or procurement
- Implement time-aware splits, baseline forecasts, seasonality review, uncertainty intervals, drift monitoring, and retraining triggers through documented architecture, policy, configuration, and change control
- Collect backtests, interval coverage, error by horizon, data-latency records, and post-change performance and compare the results with approved requirements, risk thresholds, and expected outcomes
- Exercise this failure scenario as a planned test: a model trained before a major structural change continues to produce precise but systematically wrong forecasts
Common mistakes
- Using the label time-series forecasting without defining the exact system boundary, decision, data, and people in scope
- Treating a model response or vendor claim as sufficient evidence without testing the surrounding application and operating process
- Failing to preserve backtests, interval coverage, error by horizon, data-latency records, and post-change performance in a form that lets reviewers reconstruct decisions and changes
- Testing only the normal path and never rehearsing what happens when a model trained before a major structural change continues to produce precise but systematically wrong forecasts
Certification relevance
This subject appears in or supports the following certification bodies of knowledge: