Using simulated data to develop and study diagnostic tools for data analysis is very beneficial. The user can gain insight about what happens when assumptions are violated since t...
Spectral methods for nonlinear dimensionality reduction (NLDR) impose a neighborhood graph on point data and compute eigenfunctions of a quadratic form generated from the graph. W...
Despite its state-of-the-art performance, the Data Oriented Parsing (DOP) model has been shown to suffer from biased parameter estimation, and the good performance seems more the ...
: We study mobile P2P content dissemination schemes which leverage the local dedicated caches provided by hand-held devices (e.g., smart-phones, PDAs) and opportunistic contacts be...
In this paper, a multiobjective (MO) learning approach to image feature extraction is described, where Pareto-optimal interest point (IP) detectors are synthesized using genetic p...