Forward this profile
Profile
Principal at Data, Seattle
Currently Working
Vincent Granville is a leading expert in data mining, text mining, predictive modeling, business intelligence, fraud detection, technical analysis, keyword and web analytics. Over the last ten years, he has worked in real-time credit card fraud detection with Visa, advertising analytics with CNET, A/B testing with LowerMyBills, online user experience with Wells Fargo, query intelligence with InfoSpace, click fraud detection with major search engines and large advertising clients. He founded Data Shaping Solutions in 1999 and recently partnered with Authenticlick to further develop patent-pending click fraud solutions. Vincent is a former post-doctorate of Cambridge University and University of North Carolina at Chapel Hill. He was among the finalists at the Wharton School Business Plan Competition and at the Belgian Mathematical Olympiads. --- Creation of the leading and most visited data mining meta directory and analytical job board with 150,000 visitors per year and several thousand subscribers to the email job alert. Statistical consulting until 2005. Past clients include Visa (credit card fraud detection), William Wecker Associates (statistical litigation), LowerMyBills, Wells Fargo, InfoSpace (click fraud detection). Starting in 2006, strategic partnerships with Arbita and other job ad distribution networks. Creation of the first blog and first newsgroup exclusively focused on statistical jobs. Attracted advertisers such as Amazon, Yahoo, Chase, Facebook and hedge funds in US and abroad.
Currently Working
Vincent Granville is a leading expert in data mining, text mining, predictive modeling, business intelligence, fraud detection, technical analysis, keyword and web analytics. Over the last ten years, he has worked in real-time credit card fraud detection with Visa, advertising analytics with CNET, A/B testing with LowerMyBills, online user experience with Wells Fargo, query intelligence with InfoSpace, click fraud detection with major search engines and large advertising clients. He founded Data Shaping Solutions in 1999 and recently partnered with Authenticlick to further develop patent-pending click fraud solutions. Vincent is a former post-doctorate of Cambridge University and University of North Carolina at Chapel Hill. He was among the finalists at the Wharton School Business Plan Competition and at the Belgian Mathematical Olympiads. --- Creation of the leading and most visited data mining meta directory and analytical job board with 150,000 visitors per year and several thousand subscribers to the email job alert. Statistical consulting until 2005. Past clients include Visa (credit card fraud detection), William Wecker Associates (statistical litigation), LowerMyBills, Wells Fargo, InfoSpace (click fraud detection). Starting in 2006, strategic partnerships with Arbita and other job ad distribution networks. Creation of the first blog and first newsgroup exclusively focused on statistical jobs. Attracted advertisers such as Amazon, Yahoo, Chase, Facebook and hedge funds in US and abroad.
Chief Science Officer at Authenticlick, Los Angeles
Currently,Working
Authenticlick provides advertising networks and advertisers with the most sophisticated solution for identifying click fraud and impression fraud, and more generally, for assessing traffic quality. Authenticlick combines state-of-the-art technology, deep domain expertise with years of experience in fraud detection, scoring methodology, web mining, statistical science and auditing. --- My role consists of prototyping impression / click scoring and click fraud detection algorithms, designing IP blacklists and implementing methodology to detect bogus conversions, bogus impressions, bogus clicks, as well as CTR and ad relevancy gaming both in real time and with end-of-day or end-of-week algorithms. I am responsible for developing the rule engine (including rule discovery), the machine learning platform, the validation process, as well as metric selection, user identification in various log files and standardizations procedures.
Authenticlick provides advertising networks and advertisers with the most sophisticated solution for identifying click fraud and impression fraud, and more generally, for assessing traffic quality. Authenticlick combines state-of-the-art technology, deep domain expertise with years of experience in fraud detection, scoring methodology, web mining, statistical science and auditing. --- My role consists of prototyping impression / click scoring and click fraud detection algorithms, designing IP blacklists and implementing methodology to detect bogus conversions, bogus impressions, bogus clicks, as well as CTR and ad relevancy gaming both in real time and with end-of-day or end-of-week algorithms. I am responsible for developing the rule engine (including rule discovery), the machine learning platform, the validation process, as well as metric selection, user identification in various log files and standardizations procedures.
Ph.D. [Statistics]
, University Notre Dame
[1993/December]
,
Namur, Belgium
Interests:
Internet, Finance, Business, Data Mining, Search