Feature Importance Node

F

AI/ML/Model Info

Extract per-feature importance from a Decision Tree, Random Forest or AdaBoost model

ml_feature_importanceml

Inputs3

Outputs4

Security exposure10/10

Packageml

Ratings

Scores range from 0 to 10. Higher values mean more impact, exposure, or operational weight.

SecurityAttack surface and exposure impact.

10/10High

PrivacyPotential sensitivity of processed data.

10/10High

PerformanceRuntime or resource pressure.

9/10High

GovernancePolicy, audit, or compliance impact.

10/10High

ReliabilityOperational stability considerations.

9/10High

CostExternal or compute cost impact.

10/10High

Input Pins

3

Input

Execution

exec_in

Execution trigger

Model

Struct

model

Trained tree model (Decision Tree, Random Forest or AdaBoost)

SchemaNodeMLModelNodeMLModelNodeMLModel1 fields

model_refstringrequired

Schema enforced

Feature Names

String Array

feature_names

Optional column labels in training order. Unnamed columns fall back to feature_<index>.

Default []

Output Pins

4

Done

Execution

exec_out

Activated once the importances are computed

Importance

Struct

result

Per-feature importance with leaf and depth statistics

SchemaFeatureImportanceResultFeatureImportanceResultFeatureImportanceResult8 fields

model_typestringrequired

Model kind the importances were computed from

n_featuresinteger:uintrequired

Number of feature columns covered by the report

format uintmin 0

n_treesinteger:uintrequired

Number of trees that contributed

format uintmin 0

aggregationstringrequired

How the per-tree importances were combined

featuresArray<FeatureImportance>required

One entry per feature, in column order

itemsFeatureImportancearray item

Importance of a single input feature (column of the training matrix)

indexinteger:uintrequired

Position of the feature inside the training vector

format uintmin 0

namestringrequired

Label of the feature, taken from the Feature Names input or generated as `feature_<index>`

importancenumber:doublerequired

Relative impurity decrease attributed to this feature, normalized so all entries sum to 1.0

format double

mean_impurity_decreasenumber:doublerequired

Absolute mean impurity decrease over the splits that used this feature

format double

trees_usinginteger:uintrequired

Number of trees whose feature subset contained this column

format uintmin 0

rankingArray<integer:uint>required

Column indices ordered by importance, most important first

itemsinteger:uintarray item

format uintmin 0

num_leavesinteger:uintrequired

Total number of leaves over all contributing trees

format uintmin 0

max_depthinteger:uintrequired

Deepest node depth over all contributing trees

format uintmin 0

Schema enforced

Scores

Float Array

importances

Normalized importance per feature, in column order

Top Feature

String

top_feature

Name of the most important feature

Node Info

Internal name

ml_feature_importance

Category

AI/ML/Model Info