One stop to get all your Analytic Information!
When Predictive Analytics goes to Cloud:
When Predictive Analytics goes to Cloud:
In an article published in the Communication of the ACM [1] in July 08, Brian Hayes, argues that the longstanding era of PC software is threatened by Internet-based services. Cloud computing is liberating individuals and IT departments from maintaining an elaborate infrastructure of hardware, operating systems, and low-level utilities. He points out that Amazon, Google, and IBM are deploying and testing data centers for capturing this niche market.
Unfortunately, Predictive Analytics and Data Mining software has yet to move into the cloud infrastructure and take advantage of this theoretically unlimited computing power. Even though the success of software-as-a-service (SaaS) for word processing, spreadsheeting, and on-demand customer relation management (CRM) has been highly publicized, the shift is not in the vendors’ horizon. Generally the investment in Predictive Analytics software and hardware infrastructure is considerably larger than document processing support thus raising the question of why invest in Predictive Analytics Infrastructure if an on-demand offering exists.
Recently, Zementis, Inc. has released the first cloud computing scoring engine (ADAPA, Zementis http://www.zementis.com [2]), which provides the expected cloud benefits: scalable, pay-as-you-go, reliable, as well as security through dedicated and non-shared virtual machines.
ADAPA is used through web-browser or web-service interfaces. The web- browser enables the uploading of statistical models for scoring datasets and downloading the results. The web-service mode enables a real time scoring capability to be added to any (J2EE or .net) based-system environment.
ADAPA uses predictive analytics models expressed in PMML (Predictive Modeling Markup Language), the industry standard published by the DMG (Data Mining Group). PMML handles various types of models such as decision trees, linear regression, neural networks, support vector machines, etc.
Cloud computing is not only a scalable new computing topology, but also a new paradigm for software vendors. For users, it especially brings the opportunity to run a computer-intensive application through a single web-browser interface or loosely coupled with an in-house system.
[1] Communication of the ACM, Volume 51, Issue 7 (July 2008), pp.9-11
[2] Zementis, Inc. http://www.zementis.com
In an article published in the Communication of the ACM [1] in July 08, Brian Hayes, argues that the longstanding era of PC software is threatened by Internet-based services. Cloud computing is liberating individuals and IT departments from maintaining an elaborate infrastructure of hardware, operating systems, and low-level utilities. He points out that Amazon, Google, and IBM are deploying and testing data centers for capturing this niche market.
Unfortunately, Predictive Analytics and Data Mining software has yet to move into the cloud infrastructure and take advantage of this theoretically unlimited computing power. Even though the success of software-as-a-service (SaaS) for word processing, spreadsheeting, and on-demand customer relation management (CRM) has been highly publicized, the shift is not in the vendors’ horizon. Generally the investment in Predictive Analytics software and hardware infrastructure is considerably larger than document processing support thus raising the question of why invest in Predictive Analytics Infrastructure if an on-demand offering exists.
Recently, Zementis, Inc. has released the first cloud computing scoring engine (ADAPA, Zementis http://www.zementis.com [2]), which provides the expected cloud benefits: scalable, pay-as-you-go, reliable, as well as security through dedicated and non-shared virtual machines.
ADAPA is used through web-browser or web-service interfaces. The web- browser enables the uploading of statistical models for scoring datasets and downloading the results. The web-service mode enables a real time scoring capability to be added to any (J2EE or .net) based-system environment.
ADAPA uses predictive analytics models expressed in PMML (Predictive Modeling Markup Language), the industry standard published by the DMG (Data Mining Group). PMML handles various types of models such as decision trees, linear regression, neural networks, support vector machines, etc.
Cloud computing is not only a scalable new computing topology, but also a new paradigm for software vendors. For users, it especially brings the opportunity to run a computer-intensive application through a single web-browser interface or loosely coupled with an in-house system.
[1] Communication of the ACM, Volume 51, Issue 7 (July 2008), pp.9-11
[2] Zementis, Inc. http://www.zementis.com
About
Badge
Today's SAS jobs
Visitor Profile
© 2009 Created by consultsp on Ning. Create a Ning Network!
