The Lagged Predictors Contribution

Identifying past data that most influences the signal.

At the step of identifying the model components, Smart Predict has found that previous time series have an impact on the actual values (fluctuations).

The Lagged Predictors Contribution graph shows how the signal is influenced by the recent past or far past in the case of an autoregressive component.

The lags are numbered with negative integers representing their distance in the past from the predictive forecast. Lag -1 is the point in the past just before the forecast. Lag -5 is five points in the past. The higher the absolute value, the further the point is in the past.

By default, a line chart is displayed and is read as follows:
  • The X axis displays the Lag variables sorted in reverse chronological order, recent past first.
  • The Y axis displays the Maximum Smart Variable Contribution.
Example
Let's take the following example: we have created a predictive model to forecast the ozone rate for the next 12 months.

We have obtained the following Lagged Predictors Contributions graph:

Thanks to this graph, you can identify if the ozone rate is influenced by observed values in the recent or distant past. It also gives the most important dates as well. The lags are numbered with negative integers that represent how far back in the past they are from the predictive forecasts. Smart Predict found that the previous 37 time series have an impact on the actual values. This is why the graph stops at 37. Using these lags, you can analyze how the previous values influenced the actual ones. Here you see that the lags -1, -29 and -36 are very influential.

Note
If you have choosen to get predictive forecasts per entity, you have this information for each entity.