Active learning methods seek to reduce the number of labeled examples needed to train an effective classifier, and have natural appeal in spam filtering applications where trustwo...
We address the problem of evaluating the risk of a given model accurately at minimal labeling costs. This problem occurs in situations in which risk estimates cannot be obtained f...
Christoph Sawade, Niels Landwehr, Steffen Bickel, ...
The UCSD ActiveCampus project is an exploration of wireless location-aware computing in the university setting. ActiveClass supports classroom activities such as anonymous asking ...
William G. Griswold, Patricia Shanahan, Steven W. ...
We present a region-based active contour detection algorithm for objects that exhibit relatively homogeneous photometric characteristics (e.g. smooth color or gray levels), embedd...
Stefano Soatto, Ganesh Sundaramoorthi, Anthony Yez...
This paper reports the results of a laboratory study of a proposed solution to the problem of coordinating entry into collaborative activity while protecting the privacy and contr...