CSAIL Event Calendar

Local Distance Metric Learning

Learning application-specific distance metrics from labeled
data is critical for both statistical classification
and information retrieval. Most of the earlier work in
this area has focused on finding metrics that simultaneously
optimize compactness and separability in a global
sense. Specifically, such distance metrics attempt to
keep all of the data points in each class close together
while ensuring that data points from different classes
are separated. However, particularly when classes exhibit
multimodal data distributions, these goals conflict
and thus cannot be simultaneously satisfied. This paper
proposes a Local Distance Metric (LDM) that aims to
optimize local compactness and local separability. We
present an efficient algorithm that employs eigenvector
analysis and bound optimization to learn the LDM from
training data in a probabilistic framework. We demonstrate
that LDM achieves significant improvements in
both classification and retrieval accuracy compared to
global distance learning and kernel-based KNN.

An interesting and useful application is to learn a distance metric for
selecting visually similar regions of interest indicating masses from a
reference library to assist radiologists in an interactive CAD (computer
aided diagnosis) environment. Specifically, when a suspicious mass region
is identified, we need a proper distance metric to search for similar
regions from the reference library based on the visual features. The
state-of-the-art for this task is to simply apply the Euclidean distance.
We learn a localized distance metric adaptive to each individual query.
Experiments show that, the learned adaptive localized distance metric
outperforms the Euclidean distance in this retrieval task.

See other events that are part of Biomedical Imaging and Analysis 2005/2006

See other events happening in July 2006