Tag Builder - Machine learning model tag

What is the Machine learning model tag builder (MLM)? 

The machine learning model tag builder is a TrendHub tool that enables you to deploy PMML (Predictive Model Markup Language) models converted from your previously created Python scripts. Your PMML models are used to create TrendMiner tags, that can be displayed and analysed in TrendHub. 

Note: The PMML models are a findable format when deployed from TrendMiner's Notebook tool, the platform where Python scripts can be converted into PMML models and deployed. For more information on the Notebook tool, read here.

Note: This feature is only available for user with the correct access management rights 

• Create machine learning model tags 
• Edit machine learning tags 
• Visualize machine learning tags 
• Sharing machine learning tags 
• Addition information 

How to create a machine learning tag 

1. Create a notebook and deploy (this deployment can be used as input option in the machine learning tags tool).
2. Click on the Tag Builder icon .
3. Select the Machine Learning Model option.

4. Select the PMML model of choice.
5. Map tags by selecting variables (tags)
6. Select model output.
7. Add a name and description
8. Click save.

The variables used in the machine learning model tag can be of any tag type (e.g. historian tag, imported tag, calculated tag). 

Note: Once your PMML model has been created and deployed from the Notebook tool, it is findable in the MLM tool. 

Important note: Only PMML models with 10 or less input variables will be shown in the dropdown. 

Note: Only analogue inputs and outputs (integer, float, double) are supported. 

Edit machine learning tags 

1. Open the tag builder menu. 
2. Click on the 'Saved models' hyperlink in the MLM submenu. 
3. In the work organizer select the MLM tag you want to edit. 
4. Click the edit link at the top of the menu. 
5. Edit the MLM tag. 
6. Click the 'Save' button. 

Note:  Only the creator of the MLM tag can edit and delete the tag. 

Visualize machine learning tags 

MLMT's can be visualized as any other tag using the quick or advanced tag search. Alternatively, you can: 

1. Open the tag builder menu. 
2. Click on 'Saved models' hyperlink in the MLM submenu. 
3. In the work organizer select the MLM of interest. 
4. Click the 'Add to view' button. 

Note: MLM tags are added automatically to your view after creating them for the first time. 

Sharing machine learning tags 

When a MLM is created all users can search it via the tag browser and add it to their active tag list. They will be able to use it on the chart and in searches but they will not be able to see the MLM definition by default.  

To allow other users to see the MLM the owner needs to share the Tag with them in the work organizer. After it has been shared, other users can see the MLM in their 'Shared with me' tab. By clicking it there it will open the MLM details panel where they can see the Tag definition and add it to their view. 

Additional information 

Note: Tags created using the MLMT are not immediately visible, because the tag must first be indexed in TrendMiner.  

Note: Using tags does not require a licence to use the MLMs, the licences are required for creating and editing these tags.