111 resultados para GM
Resumo:
Maximum entropy approach to classification is very well studied in applied statistics and machine learning and almost all the methods that exists in literature are discriminative in nature. In this paper, we introduce a maximum entropy classification method with feature selection for large dimensional data such as text datasets that is generative in nature. To tackle the curse of dimensionality of large data sets, we employ conditional independence assumption (Naive Bayes) and we perform feature selection simultaneously, by enforcing a `maximum discrimination' between estimated class conditional densities. For two class problems, in the proposed method, we use Jeffreys (J) divergence to discriminate the class conditional densities. To extend our method to the multi-class case, we propose a completely new approach by considering a multi-distribution divergence: we replace Jeffreys divergence by Jensen-Shannon (JS) divergence to discriminate conditional densities of multiple classes. In order to reduce computational complexity, we employ a modified Jensen-Shannon divergence (JS(GM)), based on AM-GM inequality. We show that the resulting divergence is a natural generalization of Jeffreys divergence to a multiple distributions case. As far as the theoretical justifications are concerned we show that when one intends to select the best features in a generative maximum entropy approach, maximum discrimination using J-divergence emerges naturally in binary classification. Performance and comparative study of the proposed algorithms have been demonstrated on large dimensional text and gene expression datasets that show our methods scale up very well with large dimensional datasets.
Resumo:
Synergizing graphene on silicon based nanostructures is pivotal in advancing nano-electronic device technology. A combination of molecular dynamics and density functional theory has been used to predict the electronic energy band structure and photo-emission spectrum for graphene-Si system with silicon as a substrate for graphene. The equilibrium geometry of the system after energy minimization is obtained from molecular dynamics simulations. For the stable geometry obtained, density functional theory calculations are employed to determine the energy band structure and dielectric constant of the system. Further the work function of the system which is a direct consequence of photoemission spectrum is calculated from the energy band structure using random phase approximations.
Resumo:
The influence of the flow rule on the bearing capacity of strip foundations placed on sand was investigated using a new kinematic approach of upper-bound limit analysis. The method of stress characteristics was first used to find the mechanism of the failure and to compute the stress field by using the Mohr-Coulomb yield criterion. Once the failure mechanism had been established, the kinematics of the plastic deformation was established, based on the requirements of the upper-bound limit theorem. Both associated and nonassociated plastic flows were considered, and the bearing capacity was obtained by equating the rate of external plastic work to the rate of the internal energy dissipation for both smooth and rough base foundations. The results obtained from the analysis were compared with those available from the literature. (C) 2014 American Society of Civil Engineers.
Resumo:
Patterning nanostructures on flexible substrates plays a key role in the emerging flexible electronics technology. The flexible electronic devices are inexpensive and can be conformed to any shape. The potential applications for such devices are sensors, displays, solar cells, RFID, high-density biochips, optoelectronics etc. E-beam lithography is established as a powerful tool for nanoscale fabrication, but its applicability on insulating flexible substrates is often limited because of surface charging effects. This paper presents the fabrication of nanostructures on insulating flexible substrates using low energy E-beam lithography along with metallic layers for charge dissipation. Nano Structures are patterned on different substrates of materials such as acetate and PET foils. The fabrication process parameters such as the proximity gap of exposure, the exposure dosage and developing conditions have been optimized for each substrate.
Resumo:
In this paper, we present the fabrication and characterization of Ti and Au coated hollow silicon microneedles for transdermal drug delivery applications. The hollow silicon microneedles are fabricated using isotropic etching followed by anisotropic etching to obtain a tapered tip. Silicon microneedle of 300 mu m in height, with 130 mu m outer diameter and 110 mu m inner diameter at the tip followed by 80 mu m inner diameter and 160 mu m outer diameter at the base have been fabricated. In order to improve the biocompatibility of microneedles, the fabricated microneedles were coated with Ti (500 nm) by sputtering technique followed by gold coating using electroplating. A breaking force of 225 N was obtained for the fabricated microneedles, which is 10 times higher than the skin resistive force. Hence, fabricated microneedles can easily be inserted inside the skin without breakage. The fluid flow through the microneedles was studied for different inlet pressures. A minimum inlet pressure of 0.66 kPa was required to achieve a flow rate of 50 mu l in 2 s with de-ionized water as a fluid medium. (C) 2014 Elsevier B.V. All rights reserved.
Resumo:
We report the single crystal growth of antimony doped Fe1+yTe and Fe1+yTe0.5Se0.5 (Fe1+ySbxTe1-x (x=0, 2%, 5%) and Fe1+yTe0.49Se0.49Sb0.02) by a modified horizontal Bridgman method. Growth parameters are optimized to obtain high quality single crystals. The antiferromagnetic (AFM) transition at T-N = 62.2 K which is a first order transition, shifts to lower temperature on doping in Fe1+yTe. Alternately when the chalcogen site of the ternary compound Fe1+yTe0.5Se0.5 is doped with Sb, superconductivity is preserved albeit the superconducting transition temperature (T-C) falls slightly and a concomitant reduction occurs in superconducting volume fraction. (C) 2013 Elsevier B.V. All rights reserved,
Resumo:
We report on the fabrication of polymethylmethacrylate (PMMA) nanogratings on silicon (Si) and glass substrates using electron beam lithography technique. Various aspects of proximity corrections using Monte Carlo simulation have been discussed. The fabrication process parameters such as proximity gap of exposure, exposure dosage and developing conditions have been optimized for high-density PMMA nanogratings structure on Si and glass substrates. Electron beam exposure is adjusted in such a way that PMMA acts as a negative tone resist and at the same time resolution loss due to proximity effect is minimum. Both reflection and transmission-type, nanometre period gratings have been fabricated and their diffraction characteristics are evaluated.
Resumo:
Cryosorption pump is the only possible device to pump helium, hydrogen and its isotopes in fusion environment, such as high magnetic field and high plasma temperatures. Activated carbons are known to be the most suitable adsorbent in the development of cryosorption pumps. For this purpose, the data of adsorption characteristics of activated carbons in the temperature range 4.5 K to 77 K are needed, but are not available in the literature. For obtaining the above data, a commercial micro pore analyzer operating at 77 K has been integrated with a two stage GM cryocooler, which enables the cooling of the sample temperature down to 4.5 K. A heat switch mounted between the second stage cold head and the sample chamber helps to raise the sample chamber temperature to 77 K without affecting the performance of the cryocooler. The detailed description of this system is presented elsewhere. This paper presents the results of experimental studies of adsorption isotherms measured on different types of activated carbons in the form of granules, globules, flake knitted and non-woven types in the temperature range 4.5 K to 10 K using Helium gas as the adsorbate. The above results are analyzed to obtain the pore size distributions and surface areas of the activated carbons. The effect of adhesive used for bonding the activated carbons to the panels is also studied. These results will be useful to arrive at the right choice of activated carbon to be used for the development of cryosorption pumps.
Resumo:
Milling is an energy intensive process and it is considered as one of the most energy inefficient processes. Electrical and mechanical shock loading can be used to develop a pre-treatment methodology to enhance energy efficiency of comminution and liberation of minerals. Coal and Banded Hematite Jasper (BHJ) Iron ores samples were taken for the study to know the effect of shock loading. These samples were exposed to 5 electric shocks of 300 kV using an electric shock loading device. A diaphragmless shock tube was used to produce 3 and 6 compressed air shocks of Mach number 2.12 to treat the coal and Iron ore samples. Microscopic, comminution and liberation studies were carried out to compare the effectiveness of these approaches. It was found that electric shock loading can comminute the coal samples more effectively and increases the yield of carbon by 40% at 1.6 gm/cc density over the untreated coal samples. Mechanical shock loading showed improved milling performance for both the materials and 12.90% and 8.1% reduction in the D-80 of the particles was observed during grinding for treated samples of coal and iron, respectively. Liberation of minerals in BHJ Iron ore was found unaffected due to low intensity of the mechanical shock waves and non conductivity of minerals. Compressed air based shock loading is easier to operate than electrical shock loading and it needs to be explored further to improve the energy efficacy of comminution. (C) 2014 Elsevier Ltd. All rights reserved.
Resumo:
By using the lower-bound theorem of the limit analysis in conjunction with finite elements and nonlinear optimization, bearing-capacity factors, N-c and N-gamma q, with an inclusion of pseudostatic horizontal seismic body forces, have been determined for a shallow embedded horizontal strip footing placed on sloping ground surface. The variation of N-c and N-gamma q with changes in slope angle (beta) for different values of seismic acceleration coefficient (k(h)) has been obtained. The analysis reveals that irrespective of ground inclination and the embedment depth of the footing, the factors N-c and N-gamma q decrease quite considerably with an increase in k(h). As compared with N-c, the factor N-gamma q is affected more extensively with changes in k(h) and beta. Unlike most of the results reported in literature for the seismic case, the present computational results take into account the shear resistance of soil mass above the footing level. An increase in the depth of the embedment leads to an increase in the magnitudes of both N-c and N-gamma q. (C) 2014 American Society of Civil Engineers.
Resumo:
A method is presented for determining the ultimate bearing capacity of a circular footing reinforced with a horizontal circular sheet of reinforcement placed over granular and cohesive-frictional soils. It was assumed that the reinforcement sheet could bear axial tension but not the bending moment. The analysis was performed based on the lower-bound theorem of the limit analysis in combination with finite elements and linear optimization. The present research is an extension of recent work with strip foundations reinforced with different layers of reinforcement. To incorporate the effect of the reinforcement, the efficiency factors eta(gamma) and eta(c), which need to be multiplied by the bearing capacity factors N-gamma and N-c, were established. Results were obtained for different values of the soil internal friction angle (phi). The optimal positions of the reinforcements, which would lead to a maximum improvement in the bearing capacity, were also determined. The variations of the axial tensile force in the reinforcement sheet at different radial distances from the center were also studied. The results of the analysis were compared with those available from literature. (C) 2014 American Society of Civil Engineers.
Resumo:
Reinforcing soil with fibers is a useful method for improving the strength and settlement response of soil. The soil and fiber characteristics and their interaction are some of the major factors affecting the strength of reinforced soil. The fibers are usually randomly distributed in the soil, and their orientation has a significant effect on the behavior of the reinforced soil. In the paper, a study of the effect of anisotropic distribution of fibers on the stress-strain response is presented. Based on the concept of the modified Cam clay model, an analytical model was formulated for the fiber-reinforced soil, and the effect of fiber orientation on the stress-strain behavior of soil was studied in detail. The results show that, as the inclination of fibers with the horizontal plane increased, the contribution of fibers in improving the strength of fiber-reinforced soil decreased. The effect of fibers is maximum when they are in the direction of extension, and vice versa. (C) 2014 American Society of Civil Engineers.
Resumo:
This paper presents modification of the derivation of a previously proposed constitutive model for the prediction of stress-strain behavior of municipal solid waste (MSW) incorporating different mechanisms, such as immediate compression under loading, mechanical creep, and time-dependent biodegradation effect. The model is based on critical state soil mechanics incorporating increments in volumetric strains because of elastic, plastic, creep, and biodegradation effects. The improvement introduced in this paper is the modified critical state surface and considers five variable parameters for the estimation of stress-strain behavior of MSW under different loading conditions. In addition, an expression for the strain hardening rule is derived, with considerations of time-dependent mechanical creep and biodegradation effects. The model is validated using results from experimental studies and data from published literature. The results are also compared with the predictions of the stress-strain response obtained from a well-established hyperbolic model. (c) 2014 American Society of Civil Engineers.
Resumo:
Engineering of electronic energy band structure in graphene based nanostructures has several potential applications. Substrate induced bandgap opening in graphene results several optoelectronic properties due to the inter-band transitions. Various defects like structures, including Stone-Walls and higher-order defects are observed when a graphene sheet is exfoliated from graphite and in many other growth conditions. Existence of defect in graphene based nanostructures may cause changes in optoelectronic properties. Defect engineered graphene on silicon system are considered in this paper to study the tunability of optoelectronic properties. Graphene on silicon atomic system is equilibrated using molecular dynamics simulation scheme. Based on this study, we confirm the existence of a stable super-lattice. Density functional calculations are employed to determine the energy band structure for the super-lattice. Increase in the optical energy bandgap is observed with increasing of order of the complexity in the defect structure. Optical conductivity is computed as a function of incident electromagnetic energy which is also increasing with increase in the defect order. Tunability in optoelectronic properties will be useful in understanding graphene based design of photodetectors, photodiodes and tunnelling transistors.
Resumo:
A lower-bound limit analysis formulation, by using two-dimensional finite elements, the three-dimensional Mohr-Coulomb yield criterion, and nonlinear optimization, has been given to deal with an axisymmetric geomechanics stability problem. The optimization was performed using an interior point method based on the logarithmic barrier function. The yield surface was smoothened (1) by removing the tip singularity at the apex of the pyramid in the meridian plane and (2) by eliminating the stress discontinuities at the corners of the yield hexagon in the pi-plane. The circumferential stress (sigma(theta)) need not be assumed. With the proposed methodology, for a circular footing, the bearing-capacity factors N-c, N-q, and N-gamma for different values of phi have been computed. For phi = 0, the variation of N-c with changes in the factor m, which accounts for a linear increase of cohesion with depth, has been evaluated. Failure patterns for a few cases have also been drawn. The results from the formulation provide a good match with the solutions available from the literature. (C) 2014 American Society of Civil Engineers.