Debriefing Time Series Predictive Model Results
A predictive model produces performance indicators and reports as a result of a successful training. Here is a short summary of the different components that you can use to debrief your results so you can verify the accuracy of your predictive model.
-
Is the selected performance indicator high enough to consider my predictive model robust and accurate? Check the quality of your model performance over one of the proposed performance indicators. They evaluate the "error" that would be made if the forecast was calculated in the past where the actual values are known. The Expected MAPE is the default and preferred performance indicator but you can choose any of the proposed performance indicators to evaluate your model performance. Please refer to the Time Series Forecasting Performance Indicators page for more details.
-
Which forecasts are provided by the predictive model? Have a close look at the actuals and forecasts. The Explanation tab of the report shows trends, cycles, and fluctuations in the signal, each with a description. In the Forecast tab of the report, check if there are outliers in the forecasts and detect anomalies on the actual. For more information, refer to The Predictive Forecasts, The Time Series Outliers and The Time Series Outliers (Future)
- How accurate is my predictive model? Use the Forecast vs. Actual graph to visualize the predicted values (forecast) and actual values for the data source. You can then quickly see how accurate your predictive model is, what are the outliers, and the confidence interval. For more information, refer to The Forecast vs. Actual Graph and The Time Series Outliers.
What's next?
If you are satisfied with the results of your predictive model, you can then go ahead and use it. For more information, see Saving Predictive Forecasts Generated by a Time Series Predictive Model into a Dataset or Save Predictive Forecasts Back into your Planning Model.
If you are not satisfied, try to include more historical data when preparing your data.