What does a MISSING column score mean?

A <MISSING> column score usually means that Kumo did not produce a contribution score for that column. It can occur when the selected model did not use the column or when its encoder did not produce a usable representation.

The score alone does not identify the reason. Review the column’s inferred semantic type, preprocessing configuration, and selected model plan before deciding that the column was excluded.

See Column Preprocessing, Model Plan, and Explainability.