Semi-Supervised Classification Using Sparse Gaussian Process Regression

Gaussian Processes (GPs) are promising Bayesian methods for classification and regression problems. They have also been used for semi-supervised learning tasks. In this paper, we propose a new algorithm for solving semi-supervised binary classification problem using sparse GP regression (GPR) models. It is closely related to semi-supervised learning based on support vector regression (SVR) and maximum margin clustering. The proposed algorithm is simple and easy to implement. It gives a sparse solution directly unlike the SVR based algorithm. Also, the hyperparameters are estimated easily without resorting to expensive cross-validation technique. Use of sparse GPR model helps in making the proposed algorithm scalable. Preliminary results on synthetic and real-world data sets demonstrate the efficacy of the new algorithm.

Empirical torsional potential functions from protiesn structure DATA Phi- and Psi-Potentials for Non-glycyl Amino Acid Residues

The torsional potential functions Vt(φ) and Vt(ψ) around single bonds N–Cα and Cα-C, which can be used in conformational studies of oligopeptides, polypeptides and proteins, have been derived, using crystal structure data of 22 globular proteins, fitting the observed distribution in the (φ, ψ)-plane with the value of Vtot(φ, ψ), using the Boltzmann distribution. The averaged torsional potential functions, obtained from various amino acid residues in l-configuration, are Vt(φ) = – 1.0 cos (φ + 60°); Vt(ψ) = – 0.5 cos (ψ + 60°) – 1.0 cos (2ψ + 30°) – 0.5 cos (3ψ + 30°). The dipeptide energy maps Vtot(φ, ψ) obtained using these functions, instead of the normally accepted torsional functions, were found to explain various observations, such as the absence of the left-handed alpha helix and the C7 conformation, and the relatively high density of points near the line ψ = 0°. These functions, derived from observational data on protein structures, will, it is hoped, explain various previously unexplained facts in polypeptide conformation.

An optimal parallel algorithm for sparse matrix bandwidth reduction on a linear array

A simple and efficient algorithm for the bandwidth reduction of sparse symmetric matrices is proposed. It involves column-row permutations and is well-suited to map onto the linear array topology of the SIMD architectures. The efficiency of the algorithm is compared with the other existing algorithms. The interconnectivity and the memory requirement of the linear array are discussed and the complexity of its layout area is derived. The parallel version of the algorithm mapped onto the linear array is then introduced and is explained with the help of an example. The optimality of the parallel algorithm is proved by deriving the time complexities of the algorithm on a single processor and the linear array.

Bioleaching of sulphides under applied potentials

The influence of applied DC potentials on the activity and growth of Thiobacillus ferrooxidans, as well as on the dissolution behaviour of some base metal sulphides is discussed with reference to bioleaching. Selective bioleaching of zinc from sphalerite could be achieved under an applied potential of −500 mV (saturated calomel electrode) from binary mineral mixtures containing the zinc mineral and chalcopyrite or pyrite. On the other hand, bioleaching of pyrite and chalcopyrite was found to be enhanced under positive potentials of +400 mV and +600 mV, respectively. Probable mechanisms in the electrobioleaching of sulphides are examined with respect to galvanic, microbiological and applied potential effects.

Bayesian Learning for Joint Sparse OFDM Channel Estimation and Data Detection

The impulse response of a typical wireless multipath channel can be modeled as a tapped delay line filter whose non-zero components are sparse relative to the channel delay spread. In this paper, a novel method of estimating such sparse multipath fading channels for OFDM systems is explored. In particular, Sparse Bayesian Learning (SBL) techniques are applied to jointly estimate the sparse channel and its second order statistics, and a new Bayesian Cramer-Rao bound is derived for the SBL algorithm. Further, in the context of OFDM channel estimation, an enhancement to the SBL algorithm is proposed, which uses an Expectation Maximization (EM) framework to jointly estimate the sparse channel, unknown data symbols and the second order statistics of the channel. The EM-SBL algorithm is able to recover the support as well as the channel taps more efficiently, and/or using fewer pilot symbols, than the SBL algorithm. To further improve the performance of the EM-SBL, a threshold-based pruning of the estimated second order statistics that are input to the algorithm is proposed, and its mean square error and symbol error rate performance is illustrated through Monte-Carlo simulations. Thus, the algorithms proposed in this paper are capable of obtaining efficient sparse channel estimates even in the presence of a small number of pilots.

Electrochemically deposited crystalline thin film of polyaniline on nickel for redox reactions at positive potentials

Redox reactions which occur at positive potentials such as ferrous/ferric, hydroquinone/quinone, ferrocyanide/ferricyanide etc. in aqueous acidic electrolytes cannot be studied on non-platinum metals, for example, a Ni electrode. On the contrary, these reactions occur on polyaniline (PANI) modified Ni electrodes, as evidenced from cyclic voltammetry, amperometry and steady-state polarization experiments. Under identical experimental conditions of scan rate (v) and concentration (C), the peak current density (i(p)) values of Fe2+/Fe3+ redox reaction are greater on the PANI modified Ni than on Pt. Additionally, the peak potential separation (DeltaE(p)) of the voltammogram is lesser on the PANI modified Ni. With an increase in thickness of the PANI, DeltaE(p) increases suggesting that the redox reactions tend to depart from the reversibility. Scanning electron micrographs reveal the presence of a crystalline deposit of PANI on Ni when the thickness of PANI is about 0.08 mum. However, the PANI becomes amorphous and porous at higher thickness values. Raman spectroscopy and X-ray diffraction studies corroborate the observations made out of scanning electron microscopy. Higher catalytic activity of PANI is attributed to crystalline nature of PANI on Ni. Exchange current density and standard rate constant of Fe2+/Fe(3+)redox reaction are evaluated. (C) 2002 Published by Elsevier Science B.V.

The extended Enskog operator for simple fluids with continuous potentials: single particle and collective properties

We generalized the Enskog theory originally developed for the hard-sphere fluid to fluids with continuous potentials, such as the Lennard–Jones. We derived the expression for the k and ω dependent transport coefficient matrix which enables us to calculate the transport coefficients for arbitrary length and time scales. Our results reduce to the conventional Chapman–Enskog expression in the low density limit and to the conventional k dependent Enskog theory in the hard-sphere limit. As examples, the self-diffusion of a single atom, the vibrational energy relaxation, and the activated barrier crossing dynamics problem are discussed.

Regionalization of precipitation in data sparse areas using large scale atmospheric variables - A fuzzy clustering approach

Delineation of homogeneous precipitation regions (regionalization) is necessary for investigating frequency and spatial distribution of meteorological droughts. The conventional methods of regionalization use statistics of precipitation as attributes to establish homogeneous regions. Therefore they cannot be used to form regions in ungauged areas, and they may not be useful to form meaningful regions in areas having sparse rain gauge density. Further, validation of the regions for homogeneity in precipitation is not possible, since the use of the precipitation statistics to form regions and subsequently to test the regional homogeneity is not appropriate. To alleviate this problem, an approach based on fuzzy cluster analysis is presented. It allows delineation of homogeneous precipitation regions in data sparse areas using large scale atmospheric variables (LSAV), which influence precipitation in the study area, as attributes. The LSAV, location parameters (latitude, longitude and altitude) and seasonality of precipitation are suggested as features for regionalization. The approach allows independent validation of the identified regions for homogeneity using statistics computed from the observed precipitation. Further it has the ability to form regions even in ungauged areas, owing to the use of attributes that can be reliably estimated even when no at-site precipitation data are available. The approach was applied to delineate homogeneous annual rainfall regions in India, and its effectiveness is illustrated by comparing the results with those obtained using rainfall statistics, regionalization based on hard cluster analysis, and meteorological sub-divisions in India. (C) 2011 Elsevier B.V. All rights reserved.

Bayesian framework and message passing for joint support and signal recovery of approximately sparse signals

In this paper, we develop a low-complexity message passing algorithm for joint support and signal recovery of approximately sparse signals. The problem of recovery of strictly sparse signals from noisy measurements can be viewed as a problem of recovery of approximately sparse signals from noiseless measurements, making the approach applicable to strictly sparse signal recovery from noisy measurements. The support recovery embedded in the approach makes it suitable for recovery of signals with same sparsity profiles, as in the problem of multiple measurement vectors (MMV). Simulation results show that the proposed algorithm, termed as JSSR-MP (joint support and signal recovery via message passing) algorithm, achieves performance comparable to that of sparse Bayesian learning (M-SBL) algorithm in the literature, at one order less complexity compared to the M-SBL algorithm.

Sparking potentials and ionisation coefficients in SF6

Steady-state ionisation currents under uniform field conditions have been measured in SF6 over the range 110 ÿE/pÿ1000V cmÿ1torrÿ1 with gas pressures varying from 1 to 10 torr, at 20ðC. Sparking potentials Vs were also measured for a range 1ÿpdÿ20 torr-cm. Townsend's primary ionisation (ÿ/p) and electron-attachment (ÿ/p) coefficients were found to depend on E/p only. The values of secondary-ionisation coefficient (ÿ) were also determined over the range 140ÿE/pÿ600 V cmÿ1 torrÿ1. Measurements of Vs of SF6 have shown that the deviations from Paschen's law rise up to ñ3.5% at values of pd near the Paschen minimum.

Bose-Hubbard models in confining potentials: Inhomogeneous mean-field theory

We present an extensive study of Mott insulator (MI) and superfluid (SF) shells in Bose-Hubbard (BH) models for bosons in optical lattices with harmonic traps. For this we apply the inhomogeneous mean-field theory developed by Sheshadri et al. Phys. Rev. Lett. 75, 4075 (1995)]. Our results for the BH model with one type of spinless bosons agree quantitatively with quantum Monte Carlo simulations. Our approach is numerically less intensive than such simulations, so we are able to perform calculations on experimentally realistic, large three-dimensional systems, explore a wide range of parameter values, and make direct contact with a variety of experimental measurements. We also extend our inhomogeneous mean-field theory to study BH models with harmonic traps and (a) two species of bosons or (b) spin-1 bosons. With two species of bosons, we obtain rich phase diagrams with a variety of SF and MI phases and associated shells when we include a quadratic confining potential. For the spin-1 BH model, we show, in a representative case, that the system can display alternating shells of polar SF and MI phases, and we make interesting predictions for experiments in such systems.

Gibbs Energy of Formation of Ca3Ti8Al12O37 and Phase Relations and Chemical Potentials in the System Al2O3-TiO2-CaO

The quaternary oxide in the system Al2O3-CaO-TiO2 is found to have the composition Ca3Ti8Al12O37 rather than CaTi3Al8O19 as reported in the literature. The standard Gibbs energy of formation of Ca3Ti8Al12O37 from component binary oxides is measured in the temperature range from 900 to 1250 K using a solid-state electrochemical cell incorporating single crystal CaF2 as the solid electrolyte. The results can be represented by the equation: delta G(f(0x))(0) (+/- 70)/J mol(-1) = -248474 - 15.706(T/K). Combining this information with thermodynamic data on calcium aluminates and titanates available in the literature, subsolidus phase relations in the pseudo-ternary system Al2O3-CaO-TiO2 are computed and presented as isothermal sections. The evolution of phase relations with temperature is highlighted. Chemical potential diagrams are computed at 1200 K, showing the stability domains of the various phases in the chemical potential-composition space. In each chemical potential diagram, chemical potential of one component is plotted against the cationic fraction of the other two components. The diagrams are valid at relatively high oxygen potentials where Ti is present in its four-valent state in all the oxide phases.

System Ho-Rh-O: Phase Equilibria, Chemical Potentials and Gibbs Energy of Formation of HoRhO3

The thermodynamic properties of the HoRhO3 were determined in the temperature range from 900 to 1300 K by using a solid-state electrochemical cell incorporating calcia-stabilized zirconia as the electrolyte. The standard Gibbs free energy of formation of orthorhombic perovskite HoRhO3, from Ho2O3 with C-rare earth structure and Rh2O3 with orthorhombic structure, can be expressed by the equation; Delta G(f)degrees((ox)) (+/- 78)/(J/mol) = -50535 + 3.85(T/K) Using the thermodynamic data of HoRhO3 and auxiliary data for binary oxides from the literature, the phase relations in the Ho-Rh-O system were computed at 1273 K. Thermodynamic data for intermetallic phases in the binary Ho-Rh were estimated from experimental enthalpy of formation for three compositions from the literature and Miedema's model, consistent with the phase diagram. The oxygen potential-composition diagram and three-dimensional chemical potential diagram at 1273 K, and temperature-composition diagrams at constant oxygen partial pressures were computed for the system Ho-Rh-O. The decomposition temperature of HoRhO3 is 1717(+/- 2) K in pure O-2 and 1610(+/- 2) K in air at a total pressure p(o) = 0.1 MPa.