Netflix Contest Unleashes the Power of Predictive Modeling Achievements, Demonstrating the Path to Better Understanding of Consumers and Market Dynamics

Opera Solutions, in collaboration with two other top-ranked teams ("The Ensemble") recently finished #1 on the Leaderboard in the Netflix Prize, a global analytics competition

Aug 11, 2009, 08:00 ET from Opera Solutions

NEW YORK, Aug. 11 /PRNewswire/ -- In a true photo-finish, Opera Solutions, a global management consulting firm, in partnership with two other high-ranked teams ("The Ensemble"), recently edged out a Bell Labs/Yahoo consortium in the largest and most prestigious global analytics competition ever.

The Netflix Prize competition attracted 41,000 registered teams comprised of the best and the brightest scientists and mathematicians from 186 different countries, all seeking a $1MM prize for improving U.S. online movie rental company Netflix's movie recommendation program by 10 percent or more. While the official contest winner has yet to be announced, the performance showcases Opera's leading-edge capabilities in applying advanced analytic techniques and machine learning to solve extraordinarily complex business problems.

Meeting this challenge demanded singular expertise in data management, the use of highly advanced, non-linear, adaptive, and "learning" models, and major breakthroughs in blending models' output. Many scientists were not convinced that the problem was solvable: the dataset was huge (100MM ratings, 18,000 movie titles, 500,000 customers); the customer/movie matrix was scarcely populated (only 1% contained ratings); and the ratings had to serve as both predictive and explanatory variables. Cracking the 10% barrier thus meant pushing the limits of existing modeling techniques to a significant degree.

"Opera had developed many breakthroughs in techniques and model integration in the two-and-a-half years our team spent on its own, focusing on the problem," said Arnab Gupta, CEO and founder of Opera Solutions. "In collaboration with the Ensemble, Opera Solutions' predictive modeling achievements represent a transformative opportunity to deepen companies' understanding of consumers, market dynamics, and behavior prediction."

Although Netflix offered a $1MM prize to the winner, Opera sought something different. "We brought together our team not for the money or the glory, but rather because we wanted to gain cutting-edge expertise in these techniques to apply to our clients' problems. We are already finding that these new classes of models can deliver significantly better results across our practices, in areas like credit, fraud detection, and collections, marketing and cross-selling, and valuation and trade optimization," said Jacob Spoelstra, Vice President of Analytics and the scientist who led the Opera team.

Opera's team was fielded from its three Centers of Analytic Excellence in San Diego, Shanghai, and New Delhi. They joined with Vandelay Industries and Grand Prize Team to form The Ensemble during the last month of the competition.

About Opera Solutions

Opera Solutions (www.operasolutions.com) is a global management consulting firm helping the world's premier companies quickly achieve sustained profit growth. Founded in 2004, the firm delivers results through its proprietary "Total Data, Total Insights, Rapid Action" approach. Opera's data management capabilities, advanced analytics, and commitment to end-to-end execution have unleashed billions of dollars in profit for clients in financial services, media, consumer goods, hospitality, and telecom sectors around the world. The company is applying the learnings of the competition in a wide range of areas, including credit granting, fraud and risk mitigation, marketing, pricing, and cross-selling, portfolio valuation, and trading.

SOURCE Opera Solutions



RELATED LINKS

http://www.operasolutions.com