Skip to content

Machine Learning Model

Machine Learning Models are defined as a part of Scheduling Models. This may change in the future. Therefore to create a machine learning model follow the same instructions as for creating a Scheduling Model. However for machine learning models, the supported regions are parameters, systemdata and machinelearningdata.

The following entities must be provided in machinelearningdata region for a functioning machine learning model.

  • mlmodel - Specify the feature that will be predicted using this model (output feature).
  • mlfeature - List of features and their data types. This should include both input and output features.
  • mlcsv - A SQL query that will provide training data. The columns in this query must match with the feature id's.

Example Machine Learning Model

Once the model is deployed to the database and the Dictionary Cache is refreshed a dataset record can be created for its generated PL/SQL package in Machine Learning Model page in IFS Cloud.

There are Machine Learning Interfaces that allows requesting Model Statistics and making Inference Requests for the machine learning model.