Skip Navigation

MLModelFactor

Represents a field value that has a positive or negative effect on the model’s score. This object is available in API version 53.0 and later.

Supported Calls

describeSObjects(), getDeleted(), getUpdated(), query(), retrieve()

Special Access Rules

Available with Einstein Prediction Builder and Einstein Recommendation Builder.

Fields

Field	Details
Correlation	Type double Properties Filter, Nillable, Sort Description Shows the strength of association between the variable and the outcome. The higher the correlation, the greater the association.
FactorType	Type picklist Properties Filter, Group, Nillable, Restricted picklist, Sort Description The type of factor. Possible values are: `ModelFactlet`—The field value strongly influences the outcome because the model determined that this field is always important. For example, the model can decide that the field `Industry` is always important to the outcome, regardless of its value. `ModelFactor`—The field value is important to the outcome because the field’s value is significant. For example, the model can decide that the `Annual Revenue` field value is important to the outcome because the value is above $1,000,000 or below $50,000.
Importance	Type double Properties Filter, Nillable, Sort Description Shows how much the variable influences the outcome. The higher the value, the greater the impact.
ModelId	Type reference Properties Filter, Group, Sort Description The ID of the related model. This field is a relationship field. Relationship Name Model Relationship Type Lookup Refers To MLModel
Name	Type string Properties Autonumber, Defaulted on create, Filter, idLookup, Sort Description The automatically generated ID that uniquely identifies the model.
Type	Type picklist Properties Filter, Group, Nillable, Restricted picklist, Sort Description The type of model factor. Possible values are: `And` `Basic` `Or`
Weight	Type double Properties Filter, Nillable, Sort Description Indicates how significant the field value is to the outcome or score. Model factlets tend to have higher weights than model factors.