Scoring – the process of using a model created by a data mining application like JMP to make predictions for new data – has been called the "unglamorous workhorse of data mining." Like a dark yin to the bright yang of predictive modeling, scoring plays a fundamental role in the implementation of a complete data mining life cycle. Scoring requires that the model is first adapted so that it can run where the new data is produced or stored. This process is usually a time-consuming and error-prone endeavor. In this paper, we will see how the new score code generation features in JMP 13 can assist you in extending the reach of your models while minimizing the work required to adapt them.
Nascif:
I found the presentation quite helpful. Please, would you be so kind as to share an example code or two?
Thanks & best regards,
-Matt
Hi @mattflynn,
Glad to know you found it useful!
I am working on the code examples, will post back soon.
Cheers,
Nascif
@mattflynn source code for all demos added, please check the new attachment.
As a bonus, I included an example of how to run a JMP model on a Spark cluster.
Enjoy!
Nascif
There used to be a link to a video of your presentation. I cannot longer find it. Has it been deleted?
Hi @Emolina thanks for your interest. The Discovery presentation videos were removed as a consequence of the recent JMP Community changes but we are working on bringing them back.
I will ping you when that happens. Sorry for the inconvenience.
No problem, thanks for letting me know. I watched it but I have been telling some colleagues about it. Great presentation. We have been using it for our own scoring.
Hey @Emolina, really glad to know! I added code examples for the presentation demos to this page some time ago, they might be helpful too.
Just curious (if you can share), which language have you been using as the target for scoring code generation?
We have been using Python
Thank you, the video works perfectly now
Awesome! Don't forget to check the source code for the demos. Good luck and let me know if I can be of further assistance.
A new example of a scoring application built around JMP scoring code is available at the page of our Discovery 2017 tutorial: https://community.jmp.com/t5/Discovery-Summit-2017/Sharing-the-Ultimate-Boon-A-Journey-From-Modeling...
This example shows how to use AWS Lambda services to build a scoring application capable of providing performance at scale.
Below, you'll find papers, posters and selected video clips from Discovery Summit 2016.