This paper proposes a very fast method for blindly approximating a nonlinear mapping which transforms a sum of random variables. The estimation is surprisingly good even when the basic assumption is not satisfied. We use the method for providing a good initialization for inverting post-nonlinear mixtures and Wiener systems. Experiments show that the algorithm speed is strongly improved and the ...

We propose LOCO, a distributed algorithm which solves large-scale ridge regression. LOCO randomly assigns variables to different processing units which do not communicate. Important dependencies between variables are preserved using random projections which are cheap to compute. We show that LOCO has bounded approximation error compared to the exact ridge regression solution in the fixed design...

For a graph G with weight function w on the vertices, the total distance of G is the sum over all unordered pairs of vertices x and y of w(x)w(y) times the distance between x and y. A MAD tree of G is a spanning tree with minimum total distance. We develop a linear-time algorithm to find a MAD tree of a distancehereditary graph; that is, those graphs where distances are preserved in every conne...

A method for reconstructing the complex permittivity proole of lossy di-electric objects from measured scattered far-eld data is presented. From an integral representation of the electric eld (EFIE) and applying moment method solution, an iterative reconstruction algorithm based on a conjugate gradient method is derived. A new regularized procedure is proposed based on a Markov Random Fields mo...

We show that the optimal complexity of Nesterov’s smooth first-order optimization algorithm is preserved when the gradient is only computed up to a small, uniformly bounded error. In applications of this method to semidefinite programs, this often means computing only a few dominant eigenvalues of the current iterate instead of a full matrix exponential, which significantly reduces the method’s...

The biggest limitations of learning structured predictors from big data are the computation time and the memory demands. In this paper, we propose to handle those big data problems efficiently by distributing and parallelizing the resource requirements. We present a distributed structured prediction learning algorithm for large scale models that cannot be effectively handled by a single cluster...

Over the past 20 years, a large number of epidemiological studies, particularly case-control and cohort studies, have been conducted to investigate the role of diet and the risk of developing different types of cancer. The most consistent finding so far is the association observed between consumption of vegetables and fruit and reduced risk of cancers of the digestive and respiratory tracts. Mo...

An algorithm for the minimization of energy magnetic systems is presented and applied to analysis thermal configurations a ferromagnet identify inherent structures, i.e. nearest local minima, as function temperature. Over rather narrow temperature interval, skyrmions appear reach high limit skyrmion density. In addition, performance further demonstrated in self-consistent field calculation an i...