We present a new method for learning nonlinear functions in high
dimension using semisupervised learning. Our method includes a phase
of unsupervised basis learning and a phase of supervised function
learning. The learned bases provide a set of anchor points to form a
local coordinate system, such that each data point on a high
dimensional manifold can be locally approximated by a linear
combination of its nearby anchor points, with the linear weights
offering its local-coordinate coding. We show that a high
dimensional nonlinear function can be approximated by a global linear
function with respect to this coding scheme, and the approximation
quality is ensured by the locality of such coding. The method turns
a difficult nonlinear learning problem into a simple global linear
learning problem, which overcomes some drawbacks of traditional local
learning methods. The empirical success of our method has been
demonstrated in a recent pascal image classification competition,
where the top performance was achieved by an NEC system using this
approach.
Joint work with Kai Yu at NEC Lab America
This talk is the first in our new Yahoo!-sponsored seminar Machine
Learning series. Tong is an ideal speaker, as he joined our
statistics department from Yahoo! In future weeks, we will have
speakers from several other departments with an interest in machine-
learning topics. See
http://paul.rutgers.edu/~aweinst/mlseminar.html for the current
schedule.
A pizza lunch will be served immediately following the talk!
The series is open to the entire Rutgers community. People outside
the CS department who wish to receive announcements should send their
email address to Ari Weinstein <shawarma@gmail.com> to be added to
our distribution list.
Organizers: Pavel Kuksa, Michael Littman, Vladimir Pavlovic, and Ari
Weinstein.
