Multiclass Core Vector Machine

Even though several techniques have been proposed in the literature for achieving multiclass classification using Support Vector Machine(SVM), the scalability aspect of these approaches to handle large data sets still needs much of exploration. Core Vector Machine(CVM) is a technique for scaling up a two class SVM to handle large data sets. In this paper we propose a Multiclass Core Vector Machine(MCVM). Here we formulate the multiclass SVM problem as a Quadratic Programming(QP) problem defining an SVM with vector valued output. This QP problem is then solved using the CVM technique to achieve scalability to handle large data sets. Experiments done with several large synthetic and real world data sets show that the proposed MCVM technique gives good generalization performance as that of SVM at a much lesser computational expense. Further, it is observed that MCVM scales well with the size of the data set.

A sequential dual method for structural SVMs

In many real world prediction problems the output is a structured object like a sequence or a tree or a graph. Such problems range from natural language processing to compu- tational biology or computer vision and have been tackled using algorithms, referred to as structured output learning algorithms. We consider the problem of structured classifi- cation. In the last few years, large margin classifiers like sup-port vector machines (SVMs) have shown much promise for structured output learning. The related optimization prob -lem is a convex quadratic program (QP) with a large num-ber of constraints, which makes the problem intractable for large data sets. This paper proposes a fast sequential dual method (SDM) for structural SVMs. The method makes re-peated passes over the training set and optimizes the dual variables associated with one example at a time. The use of additional heuristics makes the proposed method more efficient. We present an extensive empirical evaluation of the proposed method on several sequence learning problems.Our experiments on large data sets demonstrate that the proposed method is an order of magnitude faster than state of the art methods like cutting-plane method and stochastic gradient descent method (SGD). Further, SDM reaches steady state generalization performance faster than the SGD method. The proposed SDM is thus a useful alternative for large scale structured output learning.

Gap-Dependent Quasiparticle Dynamics and Coherent Acoustic Phonons in CaFe2As2 across Spin Density Wave Phase Transition

We report ultrafast quasiparticle (QP) dynamics and coherent acoustic phonons in undoped CaFe2As2 iron pnictide single crystals exhibiting spin-density wave (SDW) and concurrent structural phase transition at temperature T-SDW similar to 165K using femtosecond time-resolved pump-probe spectroscopy. The contributions in transient differential reflectivity arising from exponentially decaying QP relaxation and oscillatory coherent acoustic phonon mode show large variations in the vicinity of T-SDW. From the temperature-dependence of the QP recombination dynamics in the SDW phase, we evaluate a BCS-like temperature dependent charge gap with its zero-temperature value of similar to(1.6 perpendicular to 0.2)k(B)T(SDW), whereas, much above T-SDW, an electron-phonon coupling constant of similar to 0.13 has been estimated from the linear temperature-dependence of the QP relaxation time. The long-wavelength coherent acoustic phonons with typical time-period of similar to 100 ps have been analyzed in the light of propagating strain pulse model providing important results for the optical constants, sounds velocity and the elastic modulus of the crystal in the whole temperature range of 3 to 300 K.

Signatures of superconducting and pseudogap phases in ultrafast transient reflectivity of Ca(Fe0.927Co0.073)(2)As-2

We present femtosecond time-resolved pump-probe spectroscopic studies of a pseudogap (PG) along with the superconducting (SC) gap in an overdoped iron pnictide Ca(Fe0.927Co0.073)(2)As-2. It is seen that the temperature evolution of the photo-excited quasiparticle (QP) relaxation dynamics, coherently excited A(1g)-symmetric optical phonon and two acoustic phonon dynamics behave anomalously in the vicinity of the superconducting transition temperature T-c. A continuous change in the sign of the experimentally measured transient differential reflectivity Delta R/R signal at the zero time delay between the pump and probe pulses at a temperature of similar to 200K is inferred as an evidence of the emergence of the PG phase around that temperature. This behavior is independent of the pump photon energy and occurs for crystals without the spin density wave phase transition. Copyright (C) EPLA, 2014

Dependence of shear yield strain and shear transformation zone on the glass transition temperature in thin film metallic glasses

The dependence of shear yield strain, the activation energy and volume of shear transformation zone on the glass transition temperature was investigated through the analysis of statistical distributions of the first pop-in events during spherical indentation of four different thin film metallic glasses. Only the Cu-Zr metallic glass exhibits a bimodal distribution of the first pop-in loads, whereas W-Ru-B, Zr-Cu-Ni-Al and La-Co-Al metallic glasses show an unimodal distribution. Results show that shear yield strain and activation energy of shear transformation zone decrease whereas the volume of shear transformation zone increases with increasing homologous temperature, indicating that it is the activation energy rather than the volume of shear transformation zone that controls shear yield strain. (C) 2015 Elsevier B.V. All rights reserved.

On the critical thickness for non-localized to localized plastic flow transition in metallic glasses: A molecular dynamics study

Molecular dynamics simulations were employed to investigate the specimen thickness-dependent tensile behavior of a series of Cu(x)Z(100-x) (x = 20, 40, 50, 64 and 80 at%) metallic glass (MG) films, with a particular focus on the critical thickness, tc, below which non-localized plastic flow takes place. The simulation results reveal that while the transition occurs in all the alloys examined, t(c) is sensitive to the composition. We rationalize t(c) by postulating that the strain energy stored in the sample at the onset of plastic deformation has to be sufficient for the formation of shear bands. The composition-dependence of t(c) was found to correlate with the average activation energy of the atomic level plastic deformation events. (C) 2015 Elsevier Ltd. All rights reserved.