Real-Time Learning Machine; New Paradigm for Mobile Advertisers

AdTheorent's Patent-Pending Real-Time Learning Machine™ Employs Three-Stage Approach to Enabling On-the-Fly Intelligence Gathering to Enhance Mobile Ad Campaigns

For Immediate Release – New York, NY
AdTheorent, Inc., the first intelligent Real Time Bidding (RTB)-enabled mobile ad network, today announced the release of a white paper describing its patent-pending Real-Time Learning Machine(TM) (RTLM), the first real-time learning and predictive modeling platform developed specifically for mobile advertising. Authored by Dr. Saed Sayad, Chief Data Scientist for AdTheorent, the white paper, details the key differentiators of AdTheorent’s RTLM, which learns in real-time, generates data-driven predictive models “on the fly” and predicts faster than any other data mining technology, yielding demonstrable results for AdTheorent’s mobile advertisers.

AdTheorent’s RTLM application for mobile advertisers can accurately process an ever-expanding data set, including purchase data, behavioral data, psychographic data, ancillary data and social data. The company’s proprietary Traktion(TM) product also measures post-click data, delivering “beyond the click” analytics with dynamic post-click conversion data – adding to the real-time data which fuels RTLM’s predictive models.

AdTheorent’s RTLM application uses a three-stage data analysis approach which refines billions of bid requests, removes extraneous data or noise and delivers the most accurate and efficient targeting for mobile advertising campaigns.

For example, using RTLM, 100 different predictive models can be tested in less than one second, and then 80-90 percent of the irrelevant or uncorrelated variables can be eliminated ‘on the fly.’ AdTheorent’s RTLM is a breakthrough that has successfully produced 50% -500% uplift for RTLM-powered mobile advertising campaigns.

Initially, AdTheorent’s RTm Platform, which is based on RTLM and Real Time Bidding (RTB) capabilities, extracts, transforms and uploads bid requests and impression data to the cloud. Then, another extract, transfer and load (ETL) process transforms and transfers the data to a cloud-based platform for real-time analytics and Business Intelligence (BI) reporting. Finally, RTLM uses the data to enhance Click Through Rates (CTR), Cost Per Click (CPC) and Cost Per Acquisition (CPA) predictive models. This process is iterative, continuously assimilating new, real time data to deliver higher ROI for brands.

“We did not invent a new algorithm when we created the RTLM; we merely use data analysis in a different way. Traditional predictive modeling is a much more time-intensive endeavor, and is nowhere near as flexible as the approach that we have created,” said Dr. Sayad. “The real-time learning that is deployed in our system creates massive spikes on top of an existing lift in positive targeting, and this is the breakthrough that has produced the success that we are seeing in campaigns that have used this system.”

About Dr. Saed Sayad

A pioneering researcher in real-time data mining and big data analysis, Dr. Sayad has designed, developed and deployed many business and scientific applications of predictive modeling. The author of ‘An Introduction to Data Mining,’ Dr. Sayad teaches a popular graduate course in data mining at the University of Toronto, where he is an adjunct professor.

Press Inquiries

PenVine for AdTheorent
Jennifer Schenberg

Melanie Berger