We consider distributed algorithms to optimize random access multihop wireless networks in the presence of fading. Since the associated optimization problem is neither convex nor ...
In this paper, we address the matrix completion problem and propose a novel algorithm based on a smoothed rank function (SRF) approximation. Among available algorithms like FPCA a...
Belief Propagation (BP) can be very useful and efficient for performing approximate inference on graphs. But when the graph is very highly connected with strong conflicting intera...
In this paper, we present an object contour tracking approach using graph cuts based active contours (GCBAC). Our proposed algorithm does not need any a priori global shape model,...
We propose a family of novel cost-sensitive boosting methods for multi-class classification by applying the theory of gradient boosting to p-norm based cost functionals. We establ...