AutoML Importance Table should display Distance Feature names instead of DIST_X
When using the Train Using AutoML tool, the Importance Table should display Distance Feature variables using their source Feature Class name (or alias) instead of generic names such as DIST_1, DIST_2, etc.
Currently, the Importance Table generated by the Train Using AutoML geoprocessing tool:
Correctly displays raster explanatory variable names
Displays Distance Feature variables as generic labels:
DIST_1, DIST_2, DIST_3, …
When training models with multiple distance features (for example, 10 or more), it becomes difficult to interpret model importance results because users must manually map each DIST_X variable back to its original source feature class.
This manual mapping process is time-consuming, redundant, and introduces risk of human error.
Proposed Solution
Update the Importance Table output to display:
Feature Class Name or
Feature Class Alias (if available)
This would make the output more data-driven and consistent with how raster explanatory variables are handled.
Benefits
Improves model interpretability
Reduces manual post-processing work
Minimizes risk of user error
Improves usability for complex ML workflows
Creates consistency across explanatory variable types
Use Case Example
Training suitability or risk models using multiple distance-based drivers such as:
Distance to roads
Distance to rivers
Distance to urban areas
Distance to infrastructure