Seminars, Workshops and Conferences
16 August 2005: Seminar on Accelerating the Relevance Vector Machine Via data Partitioning organized by the Computer Science and Information Management (CSIM), SAT, will be held in CS Room 209, CS Building, AIT. The seminar will be conducted at 2:00 p.m. by Dr. Armin Shmilovici (http:// www.ise.bgu.ac.il/faculty/armin/), Information Systems Engineering Department, The Faculty of Engineering Sciences, Ben-Gurion University of the Negev, Israel
Abstract
The Relevance Vector Machine (RVM) is a method for training sparse
generalized linear models, and its accuracy is comparably to other machine
learning techniques. For a dataset of size N the runtime complexity of the
RVM is O(N3) and its space complexity is O(N2) which makes it too expensive
for moderately sized problems. We suggest three different algorithms which
partition the dataset into manageable chunks. Our experiments on benchmark
datasets indicate that the partition algorithms can significantly reduce
the complexity of the RVM while retaining the attractive attributes of the
original solution.
CSIM students are encouraged to attend. Others are most welcome.
For further inquiries please contact Professor Peter Haddawy /CSIM at tel. ext. 5705