Scalable non-linear Support Vector Machine using hierarchical clustering

This paper discusses a method for scaling SVM with Gaussian kernel function to handle large data sets by using a selective sampling strategy for the training set. It employs a scalable hierarchical clustering algorithm to construct cluster indexing structures of the training data in the kernel induced feature space. These are then used for selective sampling of the training data for SVM to impart scalability to the training process. Empirical studies made on real world data sets show that the proposed strategy performs well on large data sets.

Increasing ionic conductivity of polymer-sodium salt complex by addition of a non-ionic plastic crystal

Ionic conductivity and other physico-chemical properties of a soft matter composite electrolyte comprising of a polymer-sodium salt complex and a non-ionic plastic crystal are discussed here. The electrolyte under discussion comprises of polyethyleneoxide (PEO)-sodium triflate (NaCF3SO3) and succinonitrile (SN). Addition of SN to PEO-NaCF3SO3 resulted in significant enhancement in ionic conductivity. At 50% SN concentration (with respect to weight of polymer), the polymer-plastic composite electrolyte room temperature (= 25 degrees C) ionic conductivity was similar to 1.1 x 10(-4) Omega(-1) cm(-1), approximately 45 times higher than PEO-NaCF3SO3. Observations from ac-impedance spectroscopy along with X-ray diffraction, differential scanning calorimetry and Fourier transform inrared spectroscopy strongly suggest the enhancement in the composite is ionicconductivity due to enhanced ion mobility via decrease in crystallinity of PEO. The free standing composite polymer-plastic electrolytes were more compliable than PEO-NaCF3SO3 thus exhibiting no detrimental effects of succinonitrile addition on the mechanical stability of PEO-NaCF3SO3. We propose that the exploratory PEO-NaCF3SO3-SN system.discussed here will eventually be developed as a prototype electrolyte.for sodium-sulfur batteries capable of operating at ambient and.sub-ambient conditions. (C) 2010 Elsevier B.V. All rights reserved.

Non-resonant microwave absorption evidence for intrinsic Josephson coupling in Tl2Ba2CaCu2O8 single crystals

We report the results of our non-resonant microwave absorption (NRMA) studies on single crystals of Tl2Ba2CaCu2O8 (Tl 2212) which reflect the occurrence of intrinsic Josephson coupling in these crystals. We have studied the magnetic field induced microwave absorption at various temperatures from 4.2K to T-c (similar to 104K) using a standard CW EPR spectrometer (H-dc parallel to c). We observe the appearance of a characteristic feature in the NRMA signals similar to the ones observed earlier by us in Bi2Sr2CaCu2O8 (Bi 2212) starting a few degrees below T-c, which on cooling passes through a maximum in intensity before disappearing at a further low temperature. This behaviour is attributed to the appearance, strengthening and disappearance of Josephson response consequent to the temperature dependence of the viscosity of the Josephson medium between the CuO2 superconducting sheets.

Non-periodic tilings in 2-dimensions with 4, 6, 8, 10 and 12-fold symmetries

The two dimensional plane can be filled with rhombuses, so as to generate non-periodic tilings with 4, 6, 8, 10 and 12-fold symmetries. Some representative tilings constructed using the rule of inflation are shown. The numerically computed diffraction patterns for the corresponding tilings are also shown to facilitate a comparison with possible X-ray or electron diffraction pictures.

Low-Complexity Near-ML Decoding of Large Non-Orthogonal STBCs using Reactive Tabu Search

Non-orthogonal space-time block codes (STBC) with large dimensions are attractive because they can simultaneously achieve both high spectral efficiencies (same spectral efficiency as in V-BLAST for a given number of transmit antennas) as well as full transmit diversity. Decoding of non-orthogonal STBCs with large dimensions has been a challenge. In this paper, we present a reactive tabu search (RTS) based algorithm for decoding non-orthogonal STBCs from cyclic division algebras (CDA) having largedimensions. Under i.i.d fading and perfect channel state information at the receiver (CSIR), our simulation results show that RTS based decoding of 12 X 12 STBC from CDA and 4-QAM with 288 real dimensions achieves i) 10(-3) uncoded BER at an SNR of just 0.5 dB away from SISO AWGN performance, and ii) a coded BER performance close to within about 5 dB of the theoretical MIMO capacity, using rate-3/4 turbo code at a spectral efficiency of 18 bps/Hz. RTS is shown to achieve near SISO AWGN performance with less number of dimensions than with LAS algorithm (which we reported recently) at some extra complexity than LAS. We also report good BER performance of RTS when i.i.d fading and perfect CSIR assumptions are relaxed by considering a spatially correlated MIMO channel model, and by using a training based iterative RTS decoding/channel estimation scheme.

Low-Complexity Near-MAP Decoding of Large Non-Orthogonal STBCs Using PDA

Non-orthogonal space-time block codes (STBC) from cyclic division algebras (CDA) are attractive because they can simultaneously achieve both high spectral efficiencies (same spectral efficiency as in V-BLAST for a given number of transmit antennas) as well as full transmit diversity. Decoding of non-orthogonal STBCs with hundreds of dimensions has been a challenge. In this paper, we present a probabilistic data association (PDA) based algorithm for decoding non-orthogonal STBCs with large dimensions. Our simulation results show that the proposed PDA-based algorithm achieves near SISO AWGN uncoded BER as well as near-capacity coded BER (within 5 dB of the theoretical capacity) for large non-orthogonal STBCs from CDA. We study the effect of spatial correlation on the BER, and show that the performance loss due to spatial correlation can be alleviated by providing more receive spatial dimensions. We report good BER performance when a training-based iterative decoding/channel estimation is used (instead of assuming perfect channel knowledge) in channels with large coherence times. A comparison of the performances of the PDA algorithm and the likelihood ascent search (LAS) algorithm (reported in our recent work) is also presented.

Belief Propagation Based Decoding of Large Non-Orthogonal STBCs

In this paper,we present a belief propagation (BP) based algorithm for decoding non-orthogonal space-time block codes (STBC) from cyclic division algebras (CDA) having large dimensions. The proposed approachinvolves message passing on Markov random field (MRF) representation of the STBC MIMO system. Adoption of BP approach to decode non-orthogonal STBCs of large dimensions has not been reported so far. Our simulation results show that the proposed BP-based decoding achieves increasingly closer to SISO AWGN performance for increased number of dimensions. In addition, it also achieves near-capacity turbo coded BER performance; for e.g., with BP decoding of 24 x 24 STBC from CDA using BPSK (i.e.,n576 real dimensions) and rate-1/2 turbo code (i.e., 12 bps/Hz spectral efficiency), coded BER performance close to within just about 2.5 dB from the theoretical MIMO capacity is achieved.

Design of temperature-compensated reference electrodes for non-isothermal galvanic sensors

The criterion for the design of a temperature-compensated reference electrode for non-isothermal galvanic sensors is deduced from the basic flux equations of irreversible thermodynamics. It is shown that when the Seebeck coefficient of the non-isothermal cell using a solid oxygen ion-conducting electrolyte under pure oxygen is equal to the relative partial molar entropy of oxygen in the reference electrode divided by 4F, then the EMF of the non-isothermal cell is the same as that of an isothermal cell with the same electrodes operating at the higher temperature. By measuring the temperature of the melt alone and the EMF of the non-isothermal galvanic sensor, one can derive the chemical potential or the concentration of oxygen in a corrosive medium. The theory is experimentally checked using sensors for oxygen in liquid copper constructed with various metal+oxide electrodes and fully stabilised (CaO)ZrO2 as the electrolyte. To satisfy the exact condition for temperature compensation it is often necessary to have the metal or oxide as a solid solution in the reference electrode.

Direct numerical simulation of autoignition in a non-premixed, turbulent medium

In this paper, direct numerical simulation of autoignition in an initially non-premixed medium under isotropic, homogeneous, and decaying turbulence is presented. The pressure-based method developed herein is a spectral implementation of the sequential steps followed in the predictor-corrector type of algorithms; it includes the effects of density fluctuations caused by spatial inhomogeneities ill temperature and species. The velocity and pressure field are solved in the spectral space while the scalars and density field are solved in the physical space. The presented results reveal that the autoignition spots originate and evolve at locations where (1) the composition corresponds to a small range around a specific mixture fraction, and (2) the conditional scaler dissipation rate is low. A careful examination of the data obtained indicates that the autoignition spots originate in the vortex cores, and the hot gases travel outward as combustion progresses. Hence, the applicability of the transient laminar flamelet model for this problem is questioned. The dependence of autoignition characteristics on parameters such as (1) die initial eddy-turnover time and (2) the initial ratio of length scale of scalars to that of velocities are investigated. Certain implications of new results on the conditional moment closure modeling are discussed.

Non-Abelian chiral gauge theories in two dimensions

An anomalous multiflavor chiral theory, with the gauge group SU(N), is studied using non-Abelian bosonization. The theory can be made gauge invariant by introducing Wess-Zumino fields and it is particularly simple if the Jackiw-Rajaraman parameter equals 2. In the strong-coupling limit, the low-energy effective theory only contains light unconfined fermions which interact weakly.

The first D-pi-A octupolar cryptand molecule to exhibit bulk non-linearity

The cryptand derivative has H-bond mediated trigonal network structure that leads to octupolar bulk nonlinearity.

Structural alteration from non-B to B-form could reflect DNase I hypersensitivity

Preferential cleavage of active genes by DNase I has been correlated with a structurally altered conformation of DNA at the hypersensitive site in chromatin. To have a better understanding of the structural requirements for gene activation as probed by DNase I action, digestability by DNase I of synthetic polynucleotides having the ability to adopt B and non-B conformation (like Z-form) was studied which indicated a marked higher digestability of the B-form of DNA. Left handed Z form present within a natural sequence in supercoiled plasmid also showed marked resistance towards DNase I digestion. We show that alternating purine-pyrimidine sequences adopting Z-conformation exhibit DNAse I foot printing even in a protein free system. The logical deductions from the results indicate that 1) altered structure like Z-DNA is not a favourable substrate for DNase I, 2) both the ends of the alternating purine-pyrimidine insert showed hypersensitivity, 3) B-form with a minor groove of 12-13 A is a more favourable substrate for DNase I than an altered structure, 4) any structure of DNA deviating largely from B form with a capacity to flip over to the B-form are potential targets for the DNase I enzymic probes in naked DNA.