Evidence of Strong Orbital-Selective Spin-Orbital-Phonon Coupling in CrVO_{4}.

Coupling of orbital degree of freedom with a spin exchange, i.e., Kugel-Khomskii-type interaction (KK), governs a host of material properties, including colossal magnetoresistance, enhanced magnetoelectric response, and photoinduced high-temperature magnetism. In general, KK-type interactions lead to deviation in experimental observables of coupled Hamiltonian near or below the magnetic transition. Using diffraction and spectroscopy experiments, here we report anomalous changes in lattice parameters, electronic states, spin dynamics, and phonons at four times the Néel transition temperature (T_{N}) in CrVO_{4}. The temperature is significantly higher than other d-orbital compounds such as manganites and vanadates, where effects are limited to near or below T_{N}. The experimental observations are rationalized using first-principles and Green's function-based phonon and spin simulations that show unprecedentedly strong KK-type interactions via a superexchange process and an orbital-selective spin-phonon coupling coefficient at least double the magnitude previously reported for strongly coupled spin-phonon systems. Our results present an opportunity to explore the effect of KK-type interactions and spin-phonon coupling well above T_{N} and possibly bring various properties closer to application, for example, strong room-temperature magnetoelectric coupling.

Learning Gaussian-Bernoulli RBMs Using Difference of Convex Functions Optimization.

The Gaussian-Bernoulli restricted Boltzmann machine (GB-RBM) is a useful generative model that captures meaningful features from the given n -dimensional continuous data. The difficulties associated with learning GB-RBM are reported extensively in earlier studies. They indicate that the training of the GB-RBM using the current standard algorithms, namely contrastive divergence (CD) and persistent contrastive divergence (PCD), needs a carefully chosen small learning rate to avoid divergence which, in turn, results in slow learning. In this work, we alleviate such difficulties by showing that the negative log-likelihood for a GB-RBM can be expressed as a difference of convex functions if we keep the variance of the conditional distribution of visible units (given hidden unit states) and the biases of the visible units, constant. Using this, we propose a stochastic difference of convex (DC) functions programming (S-DCP) algorithm for learning the GB-RBM. We present extensive empirical studies on several benchmark data sets to validate the performance of this S-DCP algorithm. It is seen that S-DCP is better than the CD and PCD algorithms in terms of speed of learning and the quality of the generative model learned.

On Robustness of Neural Architecture Search Under Label Noise.

Neural architecture search (NAS), which aims at automatically seeking proper neural architectures given a specific task, has attracted extensive attention recently in supervised learning applications. In most real-world situations, the class labels provided in the training data would be noisy due to many reasons, such as subjective judgments, inadequate information, and random human errors. Existing work has demonstrated the adverse effects of label noise on the learning of weights of neural networks. These effects could become more critical in NAS since the architectures are not only trained with noisy labels but are also compared based on their performances on noisy validation sets. In this paper, we systematically explore the robustness of NAS under label noise. We show that label noise in the training and/or validation data can lead to various degrees of performance variations. Through empirical experiments, using robust loss functions can mitigate the performance degradation under symmetric label noise as well as under a simple model of class conditional label noise. We also provide a theoretical justification for this. Both empirical and theoretical results provide a strong argument in favor of employing the robust loss function in NAS under high-level noise.

Pressure and Temperature Dependent Structural Studies on Hollandite Type Ferrotitanate and Crystal Structure of a High Pressure Phase.

The structural stability and phase transition behavior of tetragonal (I4/m) hollandite type K2Fe2Ti6O16 have been investigated by in situ high pressure X-ray diffraction using synchrotron radiation and a diamond anvil cell as well as by variable temperature powder neutron and X-ray diffraction. The tetragonal phase is found to be stable in a wider range of temperatures, while it reversibly transforms to a monoclinic (I2/m) structure at a moderate pressure, viz. 3.6 GPa. The pressure induced phase transition occurs with only a marginal change in structural arrangements. The unit cell parameters of ambient (t) and high pressure (m) phases can be related as am ∼ at, bm ∼ ct, and cm ∼ bt. The pressure evolution of the unit cell parameters indicates anisotropic compression with ßa = ßb ≥ ßc in the tetragonal phase and becomes more anisotropic with ßa ≪ ßb < ßc in the monoclinic phase. The pressure-volume equations of state of both phases have been obtained by second order Birch-Murnaghan equations of state, and the bulk moduli are 122 and 127 GPa for tetragonal and monoclinic phases, respectively. The temperature dependent unit cell parameters show nearly isotropic expansion, with marginally higher expansion along the c-axis compared to the a- and b-axes. The tetragonal to monoclinic phase transition occurs with a reduction of unit cell volume of about 1.1% while the reduction of unit cell volume up to 6 K is only about 0.6%. The fitting of temperature dependent unit cell volume by using the Einstein model of phonons indicates the Einstein temperature is about 266(18) K.

Anomalous lattice behavior of vanadium pentaoxide (V2O5): X-ray diffraction, inelastic neutron scattering and ab initio lattice dynamics.

We present structural and dynamical studies of layered vanadium pentaoxide (V2O5). The temperature dependent X-ray diffraction measurements reveal highly anisotropic and anomalous thermal expansion from 12 K to 853 K. The results do not show any evidence of structural phase transition or decomposition of α-V2O5, contrary to the previous transmission electron microscopy (TEM) and electron energy loss spectroscopy (EELS) experiments. The inelastic neutron scattering measurements performed up to 673 K corroborate the result of our X-ray diffraction measurements. The analysis of the experimental data is carried out using ab initio lattice dynamics calculations. The important role of van der Waals dispersion and Hubbard interactions in the structure and dynamics is revealed through ab initio calculations. The calculated anisotropic thermal expansion behavior agrees well with temperature dependent X-ray diffraction. The mechanism of anisotropic thermal expansion and anisotropic linear compressibility is discussed in terms of calculated anisotropy in the Grüneisen parameters and elastic coefficients. The calculated Gibbs free energy in various phases of V2O5 is used to understand the high pressure and temperature phase diagram of the compound.

Giant Rashba effect at the topological surface of PrGe revealing antiferromagnetic spintronics.

Rashba spin-orbit splitting in the magnetic materials opens up a new perspective in the field of spintronics. Here, we report a giant Rashba spin-orbit splitting on the PrGe [010] surface in the paramagnetic phase with Rashba coefficient α R = 5 eVÅ. We find that α R can be tuned in this system as a function of temperature at different magnetic phases. Rashba type spin polarized surface states originates due to the strong hybridization between Pr 4f states with the conduction electrons. Significant changes observed in the spin polarized surface states across the magnetic transitions are due to the competition between Dzyaloshinsky-Moriya interaction and exchange interaction present in this system. Presence of Dzyaloshinsky-Moriya interaction on the topological surface give rise to Saddle point singularity which leads to electron-like and hole-like Rashba spin split bands in the [Formula: see text] and [Formula: see text] directions, respectively. Supporting evidences of Dzyaloshinsky-Moriya interaction have been obtained as anisotropic magnetoresistance with respect to field direction and first-order type hysteresis in the X-ray diffraction measurements. A giant negative magnetoresistance of 43% in the antiferromagnetic phase and tunable Rashba parameter with temperature makes this material a suitable candidate for application in the antiferromagnetic spintronic devices.

Temperature dependent structural, vibrational and magnetic properties of K3Gd5(PO4)6.

Herein we report the evolution of the crystal structure of K3Gd5(PO4)6 in the temperature range from 20 K to 1073 K, as observed from variable temperature X-ray diffraction and Raman spectroscopic studies. K3Gd5(PO4)6 has an open tunnel containing a three dimensional structure built by [Gd5(PO4)6]3- ions which in turn are formed of PO4 tetrahedra and GdOn (n = 8 and 9) polyhedra. The empty tunnels in the structure are occupied by K+ ions and maintain charge neutrality in the lattice. Evolution of unit cell parameters with temperature shows a systematic increase with temperature. The average axial thermal expansion coefficients between 20 K and 1073 K are: αa = 10.6 × 10-6 K-1, αb = 5.5 × 10-6 K-1 and αc = 16.4 × 10-6 K-1. The evolution of distortion indices of the various coordination polyhedra with temperature indicates a gradual decrease with increasing temperature, while those of Gd2O9 and K2O8 polyhedra show opposite trends. The overall anisotropy of the lattice thermal expansion is found to be controlled largely by the effect of temperature on GdOn polyhedra and their linkages. Temperature dependent Raman spectroscopic studies indicated that the intensities and wavenumbers of most of the Raman modes decrease continuously with increasing temperature. Anharmonic analyses of Raman modes indicated that the lattice, rigid translation and librational modes have larger contributions towards thermal expansion of K3Gd5(PO4)6 compared to high frequency internal modes. The temperature and field dependent magnetic measurements indicated no long range ordering down to 2 K and the observed effective magnetic moment per Gd3+ ion and the Weiss constant are 7.91 µB and 0.38 K, respectively.

Solvent transport through hard-soft segmented polymer nanocomposites.

We conducted transport studies of a common solvent (toluene) in its condensed state, through a model hard-soft segmented polyurethane-clay nanocomposite. The solvent diffusivity is observed to be non-monotonic in a functional relationship with a filler volume fraction. In stark contrast, both classical tortuous path theory based geometric calculations and free volume measurements suggest the normally expected monotonic decrease in diffusivity with increase in clay volume fraction. Large deviations between experimentally observed diffusivity coefficients and those theoretically estimated from geometric theory are also observed. However, the equilibrium swelling of a nanocomposite as indicated by the solubility coefficient did not change. To gain an insight into the solvent interaction behavior, we conducted a pre- and post swollen segmented phase analysis of pure polymers and nanocomposites. We find that in a nanocomposite, the solvent has to interact with a filler altered hard-soft segmented morphology. In the altered phase separated morphology, the spatial distribution of thermodynamically segmented hard blocks in the continuous soft matrix becomes a strong function of filler concentration. Upon solvent interaction, this spatial distribution gets reoriented due to sorption and de-clustering. The results indicate strong non-barrier influences of nanoscale fillers dispersed in phase segmented block co-polymers, affecting solvent diffusivity through them. Based on pre- and post swollen morphological observations, we postulate a possible mechanism for the non-monotonic behaviour of solvent transport for hard-soft segmented co-polymers, in which the thermodynamic phase separation is influenced by the filler.

Noise tolerance under risk minimization.

In this paper, we explore noise-tolerant learning of classifiers. We formulate the problem as follows. We assume that there is an unobservable training set that is noise free. The actual training set given to the learning algorithm is obtained from this ideal data set by corrupting the class label of each example. The probability that the class label of an example is corrupted is a function of the feature vector of the example. This would account for most kinds of noisy data one encounters in practice. We say that a learning method is noise tolerant if the classifiers learnt with noise-free data and with noisy data, both have the same classification accuracy on the noise-free data. In this paper, we analyze the noise-tolerance properties of risk minimization (under different loss functions). We show that risk minimization under 0-1 loss function has impressive noise-tolerance properties and that under squared error loss is tolerant only to uniform noise; risk minimization under other loss functions is not noise tolerant. We conclude this paper with some discussion on the implications of these theoretical results.

One-step fabrication of thermally stable TiO2/SiO2 nanocomposite microspheres by evaporation-induced self-assembly.

The evaporation-induced self-assembly of mixed colloids has been employed to synthesize microspheres of TiO(2)/SiO(2) nanocomposites. Small-angle neutron/X-ray scattering and scanning electron microscopy experiments reveal the hierarchical morphology of the microspheres. Although the internal structure of the microspheres, consisting of solely silica nanoparticles, gets significantly modified with time because of the reduction in the high specific surface area by internal coalescence, the same for the composite microspheres remains stable over an aging time of 1 year. Such temporal stability of the composite microspheres is attributed to the inhibition of coalescence of the silica nanoparticles in the presence of titania nanoparticles. X-ray diffraction and thermogravimetric results show the improved thermal stability of the composite grains against the anatase-to-rutile phase transition. Such thermal stability is attributed to the suppression of the growth of titania nanoparticles in the presence of silica nanoparticles. The UV-vis results indicate the confinement effect of the TiO(2) nanoparticles in the silica matrix. A plausible mechanism has been elucidated for the formation of microspheres with different morphology during self-assembly.

Geometric decision tree.

In this paper, we present a new algorithm for learning oblique decision trees. Most of the current decision tree algorithms rely on impurity measures to assess the goodness of hyperplanes at each node while learning a decision tree in top-down fashion. These impurity measures do not properly capture the geometric structures in the data. Motivated by this, our algorithm uses a strategy for assessing the hyperplanes in such a way that the geometric structure in the data is taken into account. At each node of the decision tree, we find the clustering hyperplanes for both the classes and use their angle bisectors as the split rule at that node. We show through empirical studies that this idea leads to small decision trees and better performance. We also present some analysis to show that the angle bisectors of clustering hyperplanes that we use as the split rules at each node are solutions of an interesting optimization problem and hence argue that this is a principled method of learning a decision tree.

Atherosclerotic plaque regression - the role of statin therapy.

Increasing knowledge of the atherosclerotic process, as well as atherosclerotic plaque composition and morphology, has lead to the identification of vulnerable plaques that lead to acute coronary syndromes. There is growing evidence for atherosclerotic plaque regression, which makes an aggressive targeted therapeutic response based on achieving plaques regression necessary in order to reduce the significant mortality and morbidity associated with coronary heart disease. This review will examine the evidence for atherosclerotic plaque regression, the important role of statins and the available imaging techniques used to investigate this condition. We will also discuss future evolving therapies and possible predictors of plaque regression, which may aid the therapeutic process.

Solvent retention, thermodynamics, rheology and small angle X-ray scattering studies on thermoreversible poly(vinylidene fluoride) gels.

Solvent retention power of poly(vinylidene fluoride) (PVDF) gels has been studied for various homologues of phthalate (aromatic diesters). The thermal stability has been examined for gels of varying morphology. Solvent evaporation, gelation, gel melting, and polymer degradation temperatures have increased with increasing aliphatic chain length of phthalates. The thermodynamics and polymer-solvent compound formations in the PVDF-phthalate gels have been explored. The weight fraction of polymer in compound has decreased with increasing aliphatic chain length. SAXS studies have confirmed the lamellar organization inside the fibrils, and interlamellar distance increases with aliphatic chain length of diesters. The scattering patterns follow the power law behavior (I(q) approximately q(-alpha)), and polymer gels consist of high-density mass (fibril), voids, and interlamellar region. Dynamic mechanical properties indicate the splintering and reformation of network structure in gels whose percolation frequency has reduced for higher aliphatic chain length phthalate. Morphology-dependent moduli have been observed, and greater mechanical strength has been verified for thicker fibrillar gels both for steady and dynamic measurements.

Low-impact scenarios may account for two-thirds of blunt traumatic aortic rupture.

BACKGROUND: Traditionally, blunt traumatic aortic rupture (BTAR) is thought to be a high-velocity injury. It was hypothesised that BTAR has a higher than suspected incidence in low-speed accidents, with unique kinematic and demographic risk factors. METHODS: Using the UK Cooperative Crash Injury Study (CCIS) framework, impact profiling was undertaken for accidents involving BTAR. Equivalence Test Speed (ETS) was the parameter used to compare crash severity within comparable impact configurations, as it is a surrogate marker reflecting the net impact forces acting on the vehicle. ETS=40 mph (the threshold used for safety testing within the EURONCAP scheme) was used to delineate low-impact blunt traumatic aortic rupture (LIBTAR) cases, which were subsequently analysed for aetiological risk factors. RESULTS: 119 fully analysed cases of aortic injury were identified from a total of 16,444 cases reported to the UK CCIS between 1998 and 2007. 79 cases (66.4%) qualified as LIBTAR. Risk factors for LIBTAR were age >60 (p<0.0001), lateral impact direction (OR 2.041, RR 1.99, p=0.003), and struck side seat position (OR 1.934, RR 1.885 p=0.101). Low-impact crash scenarios were found to represent more than 95% of UK road traffic accidents. CONCLUSION: Low-impact collisions account for two thirds of fatal aortic injuries. Age >60, lateral impacts and struck side seat position are predictive of LIBTAR. Low-impact cases were associated with minor (potentially subclinical) intimomedial injuries. Therefore, it is recommended that a higher index of suspicion of aortic injury is used in low-impact scenarios in the risk groups identified.

CD26 expression in donor stem cell harvest and its correlation with engraftment in human haematopoietic stem cell transplantation: potential predictor of early engraftment.

BACKGROUND: The efficiency of haematopoietic stem and progenitor cells (HSPCs) is important when donor cell numbers are limiting. Stable white blood cell (WBC) and platelet engraftment is crucial for the outcome of haematopoietic stem cell transplantation (HSCT). DESIGN: This article evaluates CD26/dipeptidyl peptidase-IV expression on mobilised peripheral blood stem cell (PBSC) harvest of donors and its correlation with engraftment in HSCT. We have analysed CD26 expression on cells in various gates, that is, lymphocytes, monocytes, neutrophils and all populations using flow cytometry tool. RESULTS: Ours is the first study on human mobilised PBSC harvest from cancer patients or allogeneic related donors (n = 28) to demonstrate that increased CD26 expression leads to early engraftment in transplanted cancer patients. Correlation of CD26 expression with WBC engraftment was statistically significant (lymphocyte gate: P < 0.00001; monocyte gate: P < 0.00001; neutrophil gate: P < 0.00001; all populations: P < 0.00001). CD34 expression is a known predictor of engraftment. Nevertheless, there was no correlation between CD34 and CD26 expression in these cases. CONCLUSIONS: This study has given important leads indicating that CD26 expression may be an independent predictor of engraftment. Further study with large number of patients as well as study on circulatory CD26 may add valuable information towards improving current knowledge on CD26.

A team of continuous-action learning automata for noise-tolerant learning of half-spaces.

Learning automata are adaptive decision making devices that are found useful in a variety of machine learning and pattern recognition applications. Although most learning automata methods deal with the case of finitely many actions for the automaton, there are also models of continuous-action-set learning automata (CALA). A team of such CALA can be useful in stochastic optimization problems where one has access only to noise-corrupted values of the objective function. In this paper, we present a novel formulation for noise-tolerant learning of linear classifiers using a CALA team. We consider the general case of nonuniform noise, where the probability that the class label of an example is wrong may be a function of the feature vector of the example. The objective is to learn the underlying separating hyperplane given only such noisy examples. We present an algorithm employing a team of CALA and prove, under some conditions on the class conditional densities, that the algorithm achieves noise-tolerant learning as long as the probability of wrong label for any example is less than 0.5. We also present some empirical results to illustrate the effectiveness of the algorithm.

Conditional probability-based significance tests for sequential patterns in multineuronal spike trains.

We consider the problem of detecting statistically significant sequential patterns in multineuronal spike trains. These patterns are characterized by ordered sequences of spikes from different neurons with specific delays between spikes. We have previously proposed a data-mining scheme to efficiently discover such patterns, which occur often enough in the data. Here we propose a method to determine the statistical significance of such repeating patterns. The novelty of our approach is that we use a compound null hypothesis that not only includes models of independent neurons but also models where neurons have weak dependencies. The strength of interaction among the neurons is represented in terms of certain pair-wise conditional probabilities. We specify our null hypothesis by putting an upper bound on all such conditional probabilities. We construct a probabilistic model that captures the counting process and use this to derive a test of significance for rejecting such a compound null hypothesis. The structure of our null hypothesis also allows us to rank-order different significant patterns. We illustrate the effectiveness of our approach using spike trains generated with a simulator.

Thermoreversible gelation of poly(vinylidene fluoride-co-hexafluoro propylene) in phthalates.

The thermoreversible gelation of poly(vinylidene fluoride-co-hexafluoro propylene) copolymer have been studied in a series of phthalates, Ph-(COO C(n)H(2n+1))(2) with n = 1-8. The gelation rate increases with increasing aliphatic chain length up to n = 6, and the gelation phenomena does not occur for higher n > 6. The fibrillar morphology is evident for dried gels whose dimension (both lateral and thickness) becomes shorter and thinner with increasing n. The structures of the gels formed in various phthalates have been investigated by small-angle neutron scattering and small-angle X-ray scattering techniques, suggesting sheet-like structure, where the interplanner distance increases with increasing aliphatic chain length. The scattering intensity I(q) decreases with q according to the Ornstein-Zernike model, where q = (4pi/lambda) sin theta (2theta and lambda are scattering angle and wavelength of neutron) and the correlation length, xi, assigned to the average distance between the neighboring crystallites, also increases with increasing aliphatic chain length of diesters. The detailed thermal analyses and phase diagrams of the copolymer gels have been studied in a wide range of phthalates. Further, polymer-solvent complexes leading to the formation of two distinct compounds have been reported. A systematic change of compound composition has also been observed in the whole range of phthalates studied here. On the basis of electronic structure calculation, a model has been proposed to elucidate the conformation of copolymer chain in presence of various phthalates and their complexes, which offer the cause of higher gelation rate for longer aliphatic chain length up to n = 6, no gelation phenomena occurs for n > 6, and formation of two copolymer-solvent compounds. The mechanical properties (storage modulus and viscosity) decrease with increasing aliphatic chain length of phthalates and realignment of fibrils occurs at particular frequency depending on the strength of fibrillar gels.

Statistical significance of sequential firing patterns in multi-neuronal spike trains.

Sequential firings with fixed time delays are frequently observed in simultaneous recordings from multiple neurons. Such temporal patterns are potentially indicative of underlying microcircuits and it is important to know when a repeatedly occurring pattern is statistically significant. These sequences are typically identified through correlation counts. In this paper we present a method for assessing the significance of such correlations. We specify the null hypothesis in terms of a bound on the conditional probabilities that characterize the influence of one neuron on another. This method of testing significance is more general than the currently available methods since under our null hypothesis we do not assume that the spiking processes of different neurons are independent. The structure of our null hypothesis also allows us to rank order the detected patterns. We demonstrate our method on simulated spike trains.