To Build, Buy (or Borrow): A case study on classification with “Big Data”
Details
This month we're lucky to have Brian Dalessandro, VP of Data Science at media6degrees, presenting "To Build, Buy (or Borrow): A case study on classification with “Big Data”". Pivotal has been kind enough to offer us space. Here's his abstract and bio:
Abstract: In this talk we will present our exploration of software packages for doing machine learning with “Terascale” data. We will cover the problem formulation of our classification task, look at the data that we are using and then cover why we decided to build our own big learner. We believe this is a journey many start ups may have to make, so we will try to cover the practical considerations on why we decided to build our own software rather than use what is readily available (and free). Last, we’ll cover some of the theoretical and practical elements of Stochastic Gradient Descent for Logistic Regression so that others may be able to walk away with some better footing for writing their own learning software, should they want to venture down this path.
Bio: Brian is a practicing data scientist with 12 years of modeling experience, ranging from high energy particle physics to online advertising. Brian is currently VP of Data Science at media6degrees, an online advertising firm. Over the past 4 years, Brian has been a lead researcher in developing media6degree's patent pending machine learning technology, taking the company from pre-revenue stages to supporting 100's of concurrent client campaigns. Brian's current research interests include building autonomous machine learning systems over big data architectures, causal inference and influence attribution. Brian is also a co-chair for the 2012 KDD Cup competition.
Prior to joining media6degrees, Brian was a Senior Research Analyst at Meetup.com, exploring the dynamics of group evolution and development, and before that, a credit risk modeler for American Express. Brian holds an MBA with a concentration in Statistics from NYU and a BS in both Mathematics and French Literature from Rutgers University. When not solving research problems, Brian moonlights as a semi-professional drummer and percussionist for various locally based rock bands.