During the creation of a Predictive Model rule instance, the PMML engine verifies that the PMML model that you uploaded is a well-formed XML document and adheres to its corresponding XML Schema Definition (XSD). The schema version that is used for validation is determined by the version attribute in the PMML model file.
The model file that you uploaded does not adhere to the schema definition.
Ask a PMML developer to investigate the error message that indicates the nature of the error and fix the problem. The message might contain an error code that provides hints about why the model might not have passed schema validation.
The most common errors that might be encountered are:
The value assigned to the variable is of a different type than what is defined in the PMML file. Check the corresponding model's DataDictionary element for the value's data type definition.
The value assigned to the variable is different from the allowed values. This definition should also be part of the DataDictionary section of the model.
For a categorical or ordinal variable, the allowed values are defined by using the Value element under the DataField element.
For a continuous variable, the allowed values are defined by using the Interval element under the DataField element.
There is a syntax error in declaration of the PMML element. See the PMML standard to check the correct syntax.
An element or attribute has been defined in a place that is not allowed by the PMML standard. See the PMML standard to check the correct syntax.
An element is referenced by an ID in one part of the PMML file, but the ID is either invalid or does not exist. Check if the referenced ID is present and valid.
The variable types are not consistent in the PMML file.
Attribute is not allowed in the tag.
The attribute name or ID is incorrect or uses reserved values.