Portals eNewsletters Web Seminars dataWarehouse.com DM Review Magazine
DM Review | Covering Business Intelligence, Integration & Analytics
   Covering Business Intelligence, Integration & Analytics Advanced Search
advertisement

RESOURCE PORTALS
View all Portals

WEB SEMINARS
Scheduled Events
Archived Events

RESEARCH VAULT
White Paper Library
Research Papers

CAREERZONE
View Job Listings
Post a job

Advertisement

INFORMATION CENTER
DM Review Home
Newsletters
Current Magazine Issue
Magazine Archives
Online Columnists
Ask the Experts
Industry News
Search DM Review

GENERAL RESOURCES
Bookstore
Buyer's Guide
Glossary
Industry Events Calendar
Monthly Product Guides
Software Demo Lab
Vendor Listings

DM REVIEW
About Us
Press Releases
Awards
Advertising/Media Kit
Reprints
Magazine Subscriptions
Editorial Calendar
Contact Us
Customer Service

Data Warehousing Lessons Learned:
Technology Innovation in Predictive Analytics

  Column published in DM Review Magazine
June 2004 Issue
 
  By Lou Agosta

Genetic Algorithms Break Out of the Lab

Predictive analytics has proliferated into applications to produce incremental revenue by means of customer recommendations, customer value and churn management, campaign optimization and fraud detection. When insights generated in predictive analytics are combined with collection, personalization and content management, new horizons emerge in brand equity, multichannel revenue enhancement and customer loyalty. The latest advance in predictive analytics comes in the form of genetic algorithms deployed to address issues in business analytics.

Genetic algorithms are inherently parallel, enabling large numbers of suboptimal predictive models to be invalidated simultaneously. These algorithms are surprisingly rapid in effectively searching complex, highly nonlinear, multidimensional search spaces.

Genetic algorithms mimic the parallelism of variation and selection characteristics of biological evolution. They have been used in skunk works on Wall Street to model the behavior of individual securities and even profiled on (and, yes, hyped by) media such as ABC's "Nightline." Until now, however, no vendor has applied such technology to the commercial business problems that have characterized predictive analytics and data mining. Forrester recently caught up with Doug Newell, CEO of Genalytics, as well as with companies using and considering the technology. The VC-funded startup has:

An experienced management track record. Genalytics has deep management experience in bringing innovations to market through startups in analytics, including: Newell, previously from Tessera Enterprise Systems; Whit Stockwell, previously from Kronos; and Brian Rivard, formerly of Segue Software.

Client success stories and a viable revenue model. Genalytics has many success stories and respectable client traction as evidenced by its financial services, retail and telecommunications clients. The company has been performing a series of $50,000 pilot engagements to install, build and deploy the predictive model. Predictive analytic workbenches are not shrink-wrapped applications - some assembly is required. The question is whether Genalytics is actually selling configurable software or still providing a consulting solution.

Innovative predictive analytics technology. Genalytics' main differentiator is its claim to have proprietary genetic algorithm technology to generate predictions about buying behavior, cross-selling and up-selling. Its innovation is high-performance, scalable solutions to complex problems. Genetic algorithms, first developed by John Holland in the 1970s, leverage reasoning similar to the process of variation and selection that made Darwin's theory of evolution famous. The inherent parallelism is particularly powerful in percolating a robust and nearly optimal model. Until now, this work has been of only academic interest (the kiss of death). If Genalytics continues to have success in implementing genetic algorithms in commercial contexts, it will redefine and extend the limits of what is possible with the technology. This is an exciting prospect, but it must be approached with caution and skepticism.

Genetic algorithms are not to be confused with genetic search, which has been commercially available for years. An exhaustive search of the vendors that account for 90 percent or more of the software market suggests that genetic algorithms, if used at all, have heretofore been confined to in-house, custom development in skunk works, scientific and academic context, and one-off custom projects, mostly on Wall Street. According to former Hyperparallel CEO William F. Nowacki Jr., a genetic algorithm was in the works there but never came to market prior to Yahoo!'s acquisition of the company. Finally, Thinking Machines' Darwin product was reportedly also working on commercializing a genetic algorithm (hence the name of the product), but the work was incomplete when Oracle purchased the company in June 1999.

Of course, all the usual disclaimers apply. Genalytics is a small company whose revenues are not transparent because it is privately held. One set of algorithms does not make a paradigm shift. The bottom of the chasm is littered with innovative information technology. Yet a diligent review supports the assertion that Genalytics is the first to market with this approach for commercial business applications. The behavior of algorithms is itself the subject of experimental study, and the optimization harnessed by the genetic approach and paradigm is non-random (that is, it has performance advantages). Make no mistake: There is plenty of competition to be found among the "big guys" such as IBM (DB2 Intelligent Miner), Oracle (Oracle Data Mining), NCR Teradata (Warehouse Miner), SAS Institute and SPSS. Competition notwithstanding, the quick take - and the recommendation - is that Genalytics technology is credible and worthy of further exploration. Short-list this vendor for solving high-performance problems in predictive analytics.

...............................................................................

Check out DMReview.com's resource portals for additional related content, white papers, books and other resources.

Lou Agosta, Ph.D., is a business intelligence strategist with IBM WorldWide Business Intelligence Solutions focusing on competitive dynamics. He is a former industry analyst with Giga Information Group and has served many years in the trenches as a database administrator. His book The Essential Guide to Data Warehousing is published by Prentice Hall. Please send comments and questions to Lou in care of LAgosta@acm.org.



View Full Issue View Full Magazine Issue
E-mail This Column E-Mail This Column
Printer Friendly Version Printer-Friendly Version
Request Reprints Request Reprints
Advertisement
advertisement
Site Map Terms of Use Privacy Policy
SourceMedia (c) 2005 DM Review and SourceMedia, Inc. All rights reserved.
Use, duplication, or sale of this service, or data contained herein, is strictly prohibited.