BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Machine Learning and Order Book Dynamics - Tristan Fletcher 
DTSTART:20140402T100000Z
DTEND:20140402T110000Z
UID:TALK51824@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Standard price-based techniques for predicting future movement
 s of financial assets have low performance over short predictive horizons.
  Information relating to the interaction of buyers and sellers on exchange
 s has greater prognostic ability over these time scales. With this in mind
 \, we use the volumes on a EURUSD limit order book to construct simple fea
 tures. We then take traditional models of market microstructure and incorp
 orate them into the machine learning framework using Fisher kernels. These
  Fisher kernels are combined with kernels constructed from the simple volu
 me-based features and price-based indicators using Multiple Kernel Learnin
 g. Outperformance relative to a benchmark is found for subsets of these ke
 rnels.\n
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR