Variational studies and replica symmetry breaking in the generalization problem of the binary perceptron

We analyze the average performance of a general class of learning algorithms for the nondeterministic polynomial time complete problem of rule extraction by a binary perceptron. The examples are generated by a rule implemented by a teacher network of similar architecture. A variational approach is used in trying to identify the potential energy that leads to the largest generalization in the thermodynamic limit. We restrict our search to algorithms that always satisfy the binary constraints. A replica symmetric ansatz leads to a learning algorithm which presents a phase transition in violation of an information theoretical bound. Stability analysis shows that this is due to a failure of the replica symmetric ansatz and the first step of replica symmetry breaking (RSB) is studied. The variational method does not determine a unique potential but it allows construction of a class with a unique minimum within each first order valley. Members of this class improve on the performance of Gibbs algorithm but fail to reach the Bayesian limit in the low generalization phase. They even fail to reach the performance of the best binary, an optimal clipping of the barycenter of version space. We find a trade-off between a good low performance and early onset of perfect generalization. Although the RSB may be locally stable we discuss the possibility that it fails to be the correct saddle point globally. ©2000 The American Physical Society.

Rheological behavior of binary aqueous solutions of poly(ethylene glycol) of 1500 g·mol-1 as affected by temperature and polymer concentration

The rheological behavior of poly(ethylene glycol) of 1500 g·mol -1(PEG1500) aqueous solutions with various polymer concentrations (w = 0.05, 0.10, 0.15, 0.20 and 0.25) was studied at different temperatures (T = 283.15, 288.15, 293.15, 298.15 and 303.15) K. The analyses were carried out considering shear rates ranging from (20 to 350) s-1, using a cone-and-plate rheometer under controlled stress and temperature. Classical rheological models (Newton, Bingham, Power Law, Casson, and Herschel-Bulkley) were tested. The Power Law model was shown suitable to mathematically represent the rheological behavior of these solutions. Well-adjusted empirical models were derived for consistency index variations in function of temperature (Arrhenius-type model; R2 > 0.96), polymer concentration (exponential model; R2 > 0.99) or the combination of both (R 2 > 0.99). Additionally, linear models were used to represent the variations of behavior index in the functions of temperature (R2 > 0.83) and concentration (R2 > 0.87). © 2013 American Chemical Society.

Ethanol electro-oxidation in an alkaline medium using Pd/C, Au/C and PdAu/C electrocatalysts prepared by electron beam irradiation

Carbon-supported Pd, Au and bimetallic PdAu (Pd:Au 90:10, 50:50 and 30:70 atomic ratios) electrocatalysts were prepared using electron beam irradiation. The obtained materials were characterized by energy dispersive X-ray analysis (EDX), X-ray diffraction (XRD) and transmission electron microscopy (TEM), and their catalytic activities toward ethanol electro-oxidation were evaluated in an alkaline medium using electrochemical techniques, in situ attenuated total reflectance Fourier transformed infrared spectroscopy (ATR-FTIR) analysis and a single alkaline direct ethanol fuel cell (ADEFC). EDX analyses showed that the actual Pd: Au atomic ratios were very similar to the nominal ones. X-ray diffractograms of PdAu/C electrocatalysts evidenced the presence of Pd-rich (fcc) and Au-rich (fcc) phases. TEM analysis showed a homogeneous dispersion of nanoparticles on the carbon support, with an average size in the range of 3-5 nm and broad size distributions. Cyclic voltammetry (CV) and chronoamperometry (CA) experiments revealed the superior ambient activity toward ethanol electro-oxidation of PdAu/C electrocatalysts with Pd: Au ratios of 90:10 and 50:50. In situ ATR-FTIR spectroscopy measurements have shown that the mechanism for ethanol electro-oxidation is dependent on catalyst composition, leading to different reaction products, such as acetaldehyde and acetate, depending on the number of electrons transferred. Experiments on a single ADEFC were conducted between 50 and 900 C, and the best performance of 44 mW cm-2 in 2.0molL-1 ethanol was obtained at 850C for the Pd:Au 90:10 catalysts. This superior performance is most likely associated with enhancement of ethanol adsorption on Pd, oxidation of the intermediates, the presence of gold oxide-hydroxyl species, low mean particle diameters and better distribution of particles on the support. © 2013 Elsevier Ltd. All rights reserved.

Stability and collapse of fermions in a binary dipolar boson-fermion 164Dy-161Dy mixture

We suggest a time-dependent mean-field hydrodynamic model for a binary dipolar boson-fermion mixture to study the stability and collapse of fermions in the 164Dy-161Dy mixture. The condition of stability of the dipolar mixture is illustrated in terms of phase diagrams. A collapse is induced in a disk-shaped stable binary mixture by jumping the interspecies contact interaction from repulsive to attractive by the Feshbach resonance technique. The subsequent dynamics is studied by solving the time-dependent mean-field model including three-body loss due to molecule formation in boson-fermion and boson-boson channels. Collapse and fragmentation in the fermions after subsequent explosions are illustrated. The anisotropic dipolar interaction leads to anisotropic fermionic density distribution during collapse. This study is carried out in three-dimensional space using realistic values of dipolar and contact interactions. © 2013 American Physical Society.

Exotic Bose-Einstein condensates: binary mixtures and dipolar gases

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Thermal and structural study of glasses in the binary system TeO2-Pb(PO3)(2)

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Characterization, inclusion mode, phase-solubility and in vitro release studies of inclusion binary complexes with cyclodextrins and meglumine using sulfamerazine as model drug

In order to investigate the effect on the aqueous solubility and release rate of sulfamerazine (SMR) as model drug, inclusion complexes with beta-cyclodextrin (beta CD), methyl-beta-cyclodextrin (M beta CD) and hydroxypropyl-beta-cyclodextrin (HP beta CD) and a binary system with meglumine (MEG) were developed. The formation of 1: 1 inclusion complexes of SMR with the CDs and a SMR: MEG binary system in solution and in solid state was revealed by phase solubility studies (PSS), nuclear magnetic resonance (NMR), Fourier-transform infrared spectroscopy (FT-IR), thermal analysis and X-Ray diffractometry (XRD) studies. The CDs solubilization of SMR could be improved by ionization of the drug molecule through pH adjustments. The higher apparent stability constants of SMR:CDs complexes were obtained in pH 2.00, demonstrating that CDs present more affinity for the unionized drug. The best approach for SMR solubility enhancement results from the combination of MEG and pH adjustment, with a 34-fold increment and a S-max of 54.8 mg/ml. The permeability of the drug was reduced due to the presence of beta CD, M beta CD, HP beta CD and MEG when used as solubilizers. The study then suggests interesting applications of CD or MEG complexes for modulating the release rate of SMR through semipermeable membranes.

Demixing and symmetry breaking in binary dipolar Bose-Einstein-condensate solitons

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Statics and dynamics of a binary dipolar Bose-Einstein condensate soliton

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

The NINJA-2 project: detecting and characterizing gravitational waveforms modelled using numerical binary black hole simulations

The Numerical INJection Analysis (NINJA) project is a collaborative effort between members of the numerical relativity and gravitational-wave (GW) astrophysics communities. The purpose of NINJA is to study the ability to detect GWs emitted from merging binary black holes (BBH) and recover their parameters with next-generation GW observatories. We report here on the results of the second NINJA project, NINJA-2, which employs 60 complete BBH hybrid waveforms consisting of a numerical portion modelling the late inspiral, merger, and ringdown stitched to a post-Newtonian portion modelling the early inspiral. In a 'blind injection challenge' similar to that conducted in recent Laser Interferometer Gravitational Wave Observatory (LIGO) and Virgo science runs, we added seven hybrid waveforms to two months of data recoloured to predictions of Advanced LIGO (aLIGO) and Advanced Virgo (AdV) sensitivity curves during their first observing runs. The resulting data was analysed by GW detection algorithms and 6 of the waveforms were recovered with false alarm rates smaller than 1 in a thousand years. Parameter-estimation algorithms were run on each of these waveforms to explore the ability to constrain the masses, component angular momenta and sky position of these waveforms. We find that the strong degeneracy between the mass ratio and the BHs' angular momenta will make it difficult to precisely estimate these parameters with aLIGO and AdV. We also perform a large-scale Monte Carlo study to assess the ability to recover each of the 60 hybrid waveforms with early aLIGO and AdV sensitivity curves. Our results predict that early aLIGO and AdV will have a volume-weighted average sensitive distance of 300 Mpc (1 Gpc) for 10M circle dot + 10M circle dot (50M circle dot + 50M circle dot) BBH coalescences. We demonstrate that neglecting the component angular momenta in the waveform models used in matched-filtering will result in a reduction in sensitivity for systems with large component angular momenta. This reduction is estimated to be up to similar to 15% for 50M circle dot + 50M circle dot BBH coalescences with almost maximal angular momenta aligned with the orbit when using early aLIGO and AdV sensitivity curves.

Synthesis and biological evaluation of ternary silver compounds bearing N,N-chelating ligands and thiourea: X-ray structure of [{Ag(bpy) (mu-tu)}(2)](NO3)(2) (bpy=2,2 '-bipyridine; tu = thiourea)

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Pluronics F-127/L-81 Binary Hydrogels as Drug-Delivery Systems: Influence of Physicochemical Aspects on Release Kinetics and Cytotoxicity

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Automatic classification of prostate stromal tissue in histological images using Haralick descriptors and Local Binary Patterns

In this paper we presente a classification system that uses a combination of texture features from stromal regions: Haralick features and Local Binary Patterns (LBP) in wavelet domain. The system has five steps for classification of the tissues. First, the stromal regions were detected and extracted using segmentation techniques based on thresholding and RGB colour space. Second, the Wavelet decomposition was applied in the extracted regions to obtain the Wavelet coefficients. Third, the Haralick and LBP features were extracted from the coefficients. Fourth, relevant features were selected using the ANOVA statistical method. The classication (fifth step) was performed with Radial Basis Function (RBF) networks. The system was tested in 105 prostate images, which were divided into three groups of 35 images: normal, hyperplastic and cancerous. The system performance was evaluated using the area under the ROC curve and resulted in 0.98 for normal versus cancer, 0.95 for hyperplasia versus cancer and 0.96 for normal versus hyperplasia. Our results suggest that texture features can be used as discriminators for stromal tissues prostate images. Furthermore, the system was effective to classify prostate images, specially the hyperplastic class which is the most difficult type in diagnosis and prognosis.

Poloxamer-based binary hydrogels for delivering tramadol hydrochloride: sol-gel transition studies, dissolution-release kinetics, in vitro toxicity, and pharmacological evaluation

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Local order and magnetic field effects on the electronic properties of disordered binary alloys in the quantum site percolation limit

Electronic properties of disordered binary alloys are studied via the calculation of the average Density of States (DOS) in two and three dimensions. We propose a new approximate scheme that allows for the inclusion of local order effects in finite geometries and extrapolates the behavior of infinite systems following finite-size scaling ideas. We particularly investigate the limit of the Quantum Site Percolation regime described by a tight-binding Hamiltonian. This limit was chosen to probe the role of short range order (SRO) properties under extreme conditions. The method is numerically highly efficient and asymptotically exact in important limits, predicting the correct DOS structure as a function of the SRO parameters. Magnetic field effects can also be included in our model to study the interplay of local order and the shifted quantum interference driven by the field. The average DOS is highly sensitive to changes in the SRO properties and striking effects are observed when a magnetic field is applied near the segregated regime. The new effects observed are twofold: there is a reduction of the band width and the formation of a gap in the middle of the band, both as a consequence of destructive interference of electronic paths and the loss of coherence for particular values of the magnetic field. The above phenomena are periodic in the magnetic flux. For other limits that imply strong localization, the magnetic field produces minor changes in the structure of the average DOS. © World Scientific Publishing Company.