Skip to main content
1-Visitor
November 14, 2018
Question

How to create feedback system to Build, deploy, test, and retrain a predictive machine learning mode

  • November 14, 2018
  • 2 replies
  • 3674 views

We have created data model using Thingworx. We want to improve the Accuracy of the data model by creating a feedback system in which data to be stored in database, retrieved and again send to retrain. Is there any possible way in which i can implement this?. Kindly reply to antony.joseph@maxbytetech.com. Any kind of help will be appreciated. Thank you. 

2 replies

5-Regular Member
November 14, 2018

Hello.

 

I am not aware of this sort of functionality being available in ThingWorx core. I am engaging with the ThingWorx Analytics specialists to see if that product might support adding this type of machine learning model to do what you ask.

 

Thank you,

 

-- Craig A.

vt-21-VisitorAuthor
1-Visitor
November 17, 2018

Thank you 

17-Peridot
November 14, 2018

Antony,

 

Since Maxbyte is a current client of the ThingWorx Analytics Product, would you kindly open a case?

 

The question and guidance you seek is not something as straight forward as replying to the PTC Community Post.

 

Thanks & Regards,

 

Neel

22-Sapphire I
November 14, 2018

Thingworx could potentially help, this functionality isn't out of the box.

But if the data is available in some database, Thingworx can connect to that with JDBC

Thingworx can then repost that data to the model for retrain.

I am not familiar with exactly what services Analytics exposes, so I'm not sure what has to happen when the data gets to the analytics engine.

17-Peridot
November 14, 2018

Pai,

 

All services are accessible via ThingWorx Services and APIs.

 

 

We would need more information, and discussion with Antony to better understand their use case, infrastructure, and frequency of stream, etc.

 

Regards,

 

Neel