Monitoring configuration

{

“period”: “week”, <string, Mandatory> indicates monitoring frequency running. Can be day, week, month or year “train_data” : {

“NeomarilTrainingExecution”: “1”, <string, Mandatory for model trained inside Neomaril> executionId from training promoted at Neomaril // “train_date_col”: “date”, <string, Mandatory if no field “train_date_ref” was inserted> name of the date column for the records “train_date_ref”: “2022-09-01”, <string, Mandatory if no field “train_date_col” was inserted> date the model data was acquired

} “input_cols”: [“mean_radius”, “mean_texture”] , <list[string]> name of the features columns that the monitoring will run. Need to be the same of the train data in MLFlow and the output of the pre-process function “output_cols”: [“proba”, “pred”] , <list[string]> name of the output columns that the monitoring will run. Need to be the same of the output data in MLFlow and the output of the model scoring function

}