Subscribe to DSC Newsletter

You just used your art/magic to create a great predictive model for your customer. You can't wait to turn it into business value for your customer and, of course, more revenue and business for you. But, standing in the way are all the mundane tasks and software costs of actually deploying your model. You know there must be a better, faster, and cheaper way!

We agree completely. We have faced the same problem several times ourselves. But, now there is a such a better, faster, and cheaper way. Start your own secure scoring engine in the cloud (5 minutes), deploy your model on it (1 minute), score all the data you need (for as long as you need). And, even better, pay only for as long as you need it. Sounds simple enough? Try it for yourself. Still skeptical? Let us do it for you. For free (the first time!!!). Call Zementis at 619-330-0780, ext. 2000.

About ADAPA®, Bob Nisbet, Ph.D., co-Author of the soon to be released HANDBOOK OF STATISTICAL ANALYSIS and DATA MINING APPLICATIONS, (900 pages, publication date: APR-2009, ISBN-13: 978-0-12-374765-5, Imprint: ACADEMIC PRESS), says:

ADAPA® is a truly modeling tool agnostic. Modeling tools may come and go; it makes no difference to ADAPA®. Data sets may be small to gigantic; it makes no difference to ADAPA®. It plugs into the Amazon.com elastic cloud to score data sets of any size, and (after the free trial, of course) charges a small monthly fee and according to the amount of CPU usage only. No software to buy, no installations to endure.

Models are deployed through an intuitive, Web-based management console, allowing the user to manage the deployment of decision models in a central repository. The ADAPA® product from Zementis is based on an approach developed from seemingly the opposite direction – deployment to analytical modeling. Of course, this doesn’t mean that deployment comes before modeling. Rather, this approach provides a flexible, scalable, adaptable deployment engine, which you can load with models created in any data mining package that outputs the standard Predictive Model Markup Language (which, with ADAPA®, you don’t even have to learn).

This approach is quite novel and quite powerful, because it is open-ended on the model load end (any PMML 3.2 model can input) and open-ended (almost infinitely scalable) on the deployment end, providing model predictions in a true Service Oriented Architecture (SOA) environment. Besides being data mining tool independent, on the cloud it is hardware independent as well.

Ron

Tags: Model, PMML, SOA, cloud, deployment, engine, flexible, scalable, scoring

Views: 13

Attachments:

On Data Science Central

© 2020   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service