Bayesian network (BN) parameter learning from incomplete data can be a computationally expensive task for incomplete data. Applying the EM algorithm to learn BN parameters is unfortunately susceptible to local optima and prone to premature convergence. We develop and experiment with two methods for improving EM parameter learning by using MapReduce: Age-Layered Expectation Maximization (ALEM) a...

(February 2, 1996) We investigate numerically the statistical properties of spectra of two-dimensional disordered systems by using the decimation method applied to the Anderson model. Statistics of the spacings calculated for system size up to 1024 × 1024 lattice sites exhibits a crossover between Wigner and Poisson distributions. We perform a selfcontained finite-size scaling analysis to find ...

In the framework of the Schwinger-Dyson equation and the Bethe-Salpeter equation in the improved ladder approximation, we calculate the S parameter and an analogue of the π+ − π0 mass difference ∆mπ ≡ mπ+ − mπ0 as well as the NG boson decay constant fπ on the same footing in the large Nf QCD, through the difference between the vector current correlator ΠV V and the axial-vector current correlat...

We report on a type of scaling behavior in quasiperiodically forced systems. On the parameter plane the critical point appears as a terminal point of the tori-collision bifurcation curve; its location is found numerically with high precision for two basic models, the forced supercritical circle map and the forced quadratic map. The hypothesis of universality, based on renormalization group argu...

The theory of period doublings for one-parameter families of iterated real mappings is generalized to period n-tuplings for complex mappings. An n-tupling occurs when the eigenvalue of a stable periodic orbit passes through the value ω = exp(2τπm/n) as the parameter value is changed. Each choice of m defines a different sequence of rc-tuplings, for which we construct a period ft-tupling renorma...

Ultra Thin Body (UTB) SOI are promising alternatives for extending the MOSFET scaling. However, intrinsic parameter fluctuations still remains as one of the major challenges for the ultimate scaling and integration of UTB SOI MOSFETs. In this paper, using 3D statistical numerical simulations we investigate the impact of random discrete dopants, body thickness variations and line edge roughness ...

 Background & Aims of the Study: We evaluated corrosion and scaling potential of water treated by “Takab city (Western Iran) water treatment plant (called Chahar Tagh facilities)” using field observation of water treatment plant, and study of physical and chemical parameter values of water. Materials & Methods: In this cross-sectional study, during a 10-month period (fro...

We establish an exact connection between the Choptuik scaling parameter for the three-dimensional BTZ black hole, and the imaginary part of the quasinormal frequencies for scalar perturbations. Via the AdS/CFT correspondence, this leads to an interpretation of Choptuik scaling in terms of the timescale for return to equilibrium of the dual conformal field theory.

We study the local scaling properties associated with straight line periodic orbits in homogeneous Hamiltonian systems, whose stability undergoes repeated oscillations as a function of one parameter. We give strong evidence of local scaling of the Poincaré section with exponents depending simply on the degree of homogeneity of the potential. PACS numbers: 05.45 +b To appear in Phys. Rev. Letts.

We show that imposing a certain number of vanishing moments on a scaling function (e.g., coiflets) leads to fairly small phase distortion on its associated filter bank in the neighborhood of DC. However, the phase distortion at the other frequencies can be much larger. We design a new class of real-valued, compactly supported, orthonormal, and nearly symmetric wavelets (we call them generalized...