Trainable Model¶
The Training Model page represents the Machine Learning features, the mapping between features and IFS Cloud data, and configuration options that will be passed to the Machine Learning Service for training.
While the 'General Information' section largely mirrors the content on the Trainable Models page, this page also includes additional insights in the 'Trained Model Information' section specific to each trainable model.
General Information¶
This section displays general information about the latest model training request, similar to what is shown on the Trainable Models page. The details and commands provided here are consistent with those available on that page.
Activities in Trainable Models¶
- Train Model
- Refresh Training Status
- Reset Training Status
- Refresh Model Status
- Edit/View Parameters
- Add/View Schedule
- View Background Jobs
- View Application Messages
- View Logs
Train Model¶
This will send all Machine Learning Data to the ML Service for the selected Trainable Model. The Machine Learning Data will be used to train the Trainable Model in the ML Service. Users can reset this training request by selecting the ‘Reset Training Status’ button.
Refresh Training Status¶
This will query the status of the current progress or the state of the Trainable Model’s Training Status from the ML Service and display it on the selected Trainable Model record.
Reset Training Status¶
This option allows users to reset any background jobs associated with a model training request. It becomes visible after initiating training via the 'Train Model' button.
Refresh Model Status¶
This will query the status of the Trainable Model from the ML Service and display it on the selected Trainable Model record.
Edit/View Parameters¶
This opens a dialog where the parameters for the Trainable Model can be modified or viewed.
Parameters can only be modified when the Trainable Model is not currently being trained.
Add/View Schedule¶
If the Trainable Model does not have a schedule already, it's possible to use the Add Schedule option to add a schedule to the Trainable Model.
Once a schedule has been added, the schedule can be viewed and modified using the View Schedule option. This will navigate to a detailed view of the schedule where the parameters can be edited and the schedule can be activated as per the requirement.
Note: The Next Execution Date of the model will not be displayed unless a schedule is activated. Activating a schedule means choosing to re-train the Trainable Model on available data from IFS Cloud at a pre-scheduled date/time.
View Background Jobs¶
This allows users to view the history of the Model Training Jobs that were run.
More details related to a background job can be viewed by selecting Show Details for the background job.
Read more about Background Jobs.
View Application Messages¶
This allows viewing the message requests and responses to and from the Machine Learning Service.
Read more about Application Messages.
View Logs¶
This will navigate to the Machine Learning Logs page and display logs for the selected Trainable Model.
Read more about Machine Learning Logs.
Trained Model Information¶
This section contains the historical logs of all training requests, including relevant details for each request. Further, Model Information and AI Explanations will be shown for each successful training request.
Activities in Trainable Model¶
Model Information¶
This enables users to view detailed information about the machine learning model through the following options. - Model Statistics – Statistical information about the trained model. - Dataset Profile – Statistical information about the dataset that is used for training the model. - Alert Information - Warnings and errors of the trained model.
Download Model Information¶
This feature allows users to download the model information for each successfully completed training. It combines details such as model statistics, dataset profiles, and alert information, providing a comprehensive summary for further analysis.
AI Explanation¶
This opens a dialog displaying AI explanations related to the performance and behavior of the machine learning model. It offers an overview of the key factors influencing model performance and highlights the contribution of each feature to the model's final predictions.