Predictive Intelligence Similarity definition Implementation Process

[Vivek Verma]

I implemented the PI similarity definition on Dev Instance. And added the solution to the current updateSet and deployed it to a higher instance.

But I can see Similarity Definition and ML solutions have been moved successfully, but Training REquest Schedules are missing. So do I need to forcefully move the Training Request Schedules to a higher instance?

Dev Instance:

Higher Instance:

Or do I need to click on update&retrain again?

[Brian Bakker · Accepted solution]

@Vivek Verma 

When you click on Update & Retrain on the target instance, it will create the Training Request Schedule records.

[Brian Bakker · Accepted solution]

@Vivek Verma 

When you click on Update & Retrain on the target instance, it will create the Training Request Schedule records.

[Vivek Verma]

Thank you @Brian Bakker Brian. So I need to manually update&train the definition on Prod as well? 

[Brian Bakker · Accepted solution]

@Vivek Verma 

We would also recommend to manually update & train the solution definition on PROD, as this will train the solution on the latest production data, unless you had cured your data for machine learning. ML training has little impact on the instance, as this is done on our ML Trainers in our data centers.

[Vivek Verma]

Thank you... @Brian Bakker