Dr. Matthew Peters is the Chief Data Scientist at SEOmoz, where he manages a team broadly responsible for all of SEOmoz's metrics. In his two years at the company he has developed several different machine learning algorithms including: an update to Page Authority and Domain Authority; Social Authority, a measure of influence on Twitter integrated into Followerwonk; Feed Authority, a ranking of RSS feeds based on their readership levels; and Dragnet, a content extraction/de-chroming algorithm that will be published at the 22nd International World Wide Web Conference.
In addition, he lead the SEOmoz Ranking Factors 2011 project and is currently leading this year's project. He has also worked on algorithmic web spam detection, CrawlRank an internal metric used for intelligent scheduling inside the Mozscape crawlers, and customer churn and site usage analysis to inform marketing and product decisions.
Prior to SEOmoz, Matt worked in the finance industry, initially at WaMu and Chase, and later as a consultant to Fannie Mae. At WaMu he built mortgage prepayment models and after the financial meltdown was instrumental in developing a loan level prepayment/credit model used at Chase to manage the combined WaMu-Chase mortgage portfolio. He took this experience to Fannie Mae where he consulted for their Loss Forecasting and Analysis team, the group responsible for forecasting credit losses to the US Treasury.
Previous to the finance industry, Matt worked in climate sciences, both as a postdoctoral fellow at Harvard's Earth and Planetary Sciences and as a Ph.D. student at the University of Washington. He studied the impact of clouds in the tropics on global climate and estimated their feedback effect in climate change. During that time, he published a host of peer-reviewed journal articles, conference proceedings and spoke at academic conferences. Matt has a Ph.D. in Applied Mathematics from the University of Washington.