The Machine Learning Forecast Quality layer displays the quality of the real time forecast results provided by Machine Learning Forecast (MLF) engine, compared to the real measurements retrieved from the field.
On the map you can see the relative error of forecast measurements vs current real measurements.
The associated KPI measures this relative error. An increasing KPI means an increasing relative error, that's means a worse condition traffic forecast.
The KPI can be focused on speed or flow, according with your selection in LAYER OPTIONS panel.
LAYER INFO
| Info type | Description |
|---|---|
|
|
If you have selected flow in the LAYER OPTIONS panel, on the map you see a KPI based on Geoffry E. Havers (GEH) formula. If the KPI colour is GREEN, the relative error between forecast measurements vs current real measurements, is low or 0 (for an ideal forecast, GEH=0). The relative error increase proceeding from GREEN to YELLOW, and RED. |
|
|
If you have selected speed in the LAYER OPTIONS panel, on the map you see a KPI based on Relative Absolute Error (RAE). RAE is coincident with SFQI index (see → Forecast Quality Indicators (FQI)) If the KPI colour is GREEN, the relative error between forecast measurements vs current real measurements, is low or 0 (for an ideal forecast, RAE=0). The relative error increase proceeding from GREEN to YELLOW, ORANGE, and RED. |
LAYER OPTIONS
| Option type | Description |
|---|---|
| Forecast |
Choose the time interval (in minutes) for your forecast. Tip: The duration of the interval is set according to the configuration of your Optima instance. The layer displays the traffic forecast data according to the selected value. |
| Value Type |
You can get forecast related to:
|
Machine Learning Forecast Quality layer attributes
- Click on a street (or link) to open the pop-up window with detailed information about the attributes.
The selected street (or link) is colored with a purple bold line.
Other common operations
- Select the layer in the ACTIVE LAYERS list.
- Click the
List icon.
A pop-up window with a list of items opens. Every item represents a street (or link), characterized by a set of attributes.
Other operations are available for the layer (→ Operations on layers).