Search results for: 2D function approximation.

Authors: Fethi Soltani, Adel Almarashi, Idir Mechai

Abstract:

Tikhonov regularization and reproducing kernels are the most popular approaches to solve ill-posed problems in computational mathematics and applications. And the Fourier multiplier operators are an essential tool to extend some known linear transforms in Euclidean Fourier analysis, as: Weierstrass transform, Poisson integral, Hilbert transform, Riesz transforms, Bochner-Riesz mean operators, partial Fourier integral, Riesz potential, Bessel potential, etc. Using the theory of reproducing kernels, we construct a simple and efficient representations for some class of Fourier multiplier operators Tm on the Paley-Wiener space Hh. In addition, we give an error estimate formula for the approximation and obtain some convergence results as the parameters and the independent variables approaches zero. Furthermore, using numerical quadrature integration rules to compute single and multiple integrals, we give numerical examples and we write explicitly the extremal function and the corresponding Fourier multiplier operators.

Keywords: Fourier multiplier operators, Gauss-Kronrod method of integration, Paley-Wiener space, Tikhonov regularization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1467

Authors: U.K. Saha, L.K. Arora

Abstract:

The multiindex Mittag-Leffler (M-L) function and the multiindex Dzrbashjan-Gelfond-Leontiev (D-G-L) differentiation and integration play a very pivotal role in the theory and applications of generalized fractional calculus. The object of this paper is to investigate the relations that exist between the Riemann-Liouville fractional calculus and multiindex Dzrbashjan-Gelfond-Leontiev differentiation and integration with multiindex Mittag-Leffler function.

Keywords: Multiindex Mittag-Leffler function, Multiindex Dzrbashjan-Gelfond-Leontiev differentiation and integration, Riemann-Liouville fractional integrals and derivatives.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1483

Authors: Byung-Joo Oh

Abstract:

This paper describes a method to improve the robustness of a face recognition system based on the combination of two compensating classifiers. The face images are preprocessed by the appearance-based statistical approaches such as Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). LDA features of the face image are taken as the input of the Radial Basis Function Network (RBFN). The proposed approach has been tested on the ORL database. The experimental results show that the LDA+RBFN algorithm has achieved a recognition rate of 93.5%

Keywords: Face recognition, linear discriminant analysis, radial basis function network.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2069

Authors: M. Govindarajan, R. M.Chandrasekaran

Abstract:

Text data mining is a process of exploratory data analysis. Classification maps data into predefined groups or classes. It is often referred to as supervised learning because the classes are determined before examining the data. This paper describes proposed radial basis function Classifier that performs comparative crossvalidation for existing radial basis function Classifier. The feasibility and the benefits of the proposed approach are demonstrated by means of data mining problem: direct Marketing. Direct marketing has become an important application field of data mining. Comparative Cross-validation involves estimation of accuracy by either stratified k-fold cross-validation or equivalent repeated random subsampling. While the proposed method may have high bias; its performance (accuracy estimation in our case) may be poor due to high variance. Thus the accuracy with proposed radial basis function Classifier was less than with the existing radial basis function Classifier. However there is smaller the improvement in runtime and larger improvement in precision and recall. In the proposed method Classification accuracy and prediction accuracy are determined where the prediction accuracy is comparatively high.

Keywords: Text Data Mining, Comparative Cross-validation, Radial Basis Function, runtime, accuracy.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1508

Authors: Satyendra Nath Mandal, J. Pal Choudhury, Dilip De, S. R. Bhadra Chaudhuri

Abstract:

The selection for plantation of a particular type of mustard plant depending on its productivity (pod yield) at the stage of maturity. The growth of mustard plant dependent on some parameters of that plant, these are shoot length, number of leaves, number of roots and roots length etc. As the plant is growing, some leaves may be fall down and some new leaves may come, so it can not gives the idea to develop the relationship with the seeds weight at mature stage of that plant. It is not possible to find the number of roots and root length of mustard plant at growing stage that will be harmful of this plant as roots goes deeper to deeper inside the land. Only the value of shoot length which increases in course of time can be measured at different time instances. Weather parameters are maximum and minimum humidity, rain fall, maximum and minimum temperature may effect the growth of the plant. The parameters of pollution, water, soil, distance and crop management may be dominant factors of growth of plant and its productivity. Considering all parameters, the growth of the plant is very uncertain, fuzzy environment can be considered for the prediction of shoot length at maturity of the plant. Fuzzification plays a greater role for fuzzification of data, which is based on certain membership functions. Here an effort has been made to fuzzify the original data based on gaussian function, triangular function, s-function, Trapezoidal and L –function. After that all fuzzified data are defuzzified to get normal form. Finally the error analysis (calculation of forecasting error and average error) indicates the membership function appropriate for fuzzification of data and use to predict the shoot length at maturity. The result is also verified using residual (Absolute Residual, Maximum of Absolute Residual, Mean Absolute Residual, Mean of Mean Absolute Residual, Median of Absolute Residual and Standard Deviation) analysis.

Keywords: Fuzzification, defuzzification, gaussian function, triangular function, trapezoidal function, s-function, , membership function, residual analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2272

Authors: S. Choodoung, U. Smutkupt

Abstract:

This study systemizes processes and methods in wooden furniture design that contains uniqueness in function and aesthetics. The study was done by research and analysis for designer-s consideration factors that affect function and production. Therefore, the study result indicates that such factors are design process (planning for design, product specifications, concept design, product architecture, industrial design, production), design evaluation as well as wooden furniture design dependent factors i.e. art (art style; furniture history, form), functionality (the strength and durability, area place, using), material (appropriate to function, wood mechanical properties), joints, cost, safety, and social responsibility. Specifically, all aforementioned factors affect good design. Resulting from direct experience gained through user-s usage, the designer must design the wooden furniture systemically and effectively. As a result, this study selected dinning armchair as a case study with all involving factors and all design process stated in this study.

Keywords: Furniture Design, Function Design, Aesthetic, Wooden Furniture.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 9913

Authors: Hazem J. Smadi

Abstract:

FMEA has been used for several years and proved its efficiency for system’s risk analysis due to failures. Risk priority number found in FMEA is used to rank failure modes that may occur in a system. There are some guidelines in the literature to assign the values of FMEA components known as Severity, Occurrence and Detection. This paper propose a method to assign the value for occurrence in more realistic manner representing the state of the system under study rather than depending totally on the experience of the analyst. This method uses the hazard function of a system to determine the value of occurrence depending on the behavior of the hazard being constant, increasing or decreasing.

Keywords: FMEA, Hazard Function, Risk Priority Number.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3478

Authors: S. H. Yahaya, J. M. Ali, T.A. Abdullah

Abstract:

The exploration of this paper will focus on the Cshaped transition curve. This curve is designed by using the concept of circle to circle where one circle lies inside other. The degree of smoothness employed is curvature continuity. The function used in designing the C-curve is Bézier-like cubic function. This function has a low degree, flexible for the interactive design of curves and surfaces and has a shape parameter. The shape parameter is used to control the C-shape curve. Once the C-shaped curve design is completed, this curve will be applied to design spur gear tooth. After the tooth design procedure is finished, the design will be analyzed by using Finite Element Analysis (FEA). This analysis is used to find out the applicability of the tooth design and the gear material that chosen. In this research, Cast Iron 4.5 % Carbon, ASTM A-48 is selected as a gear material.

Keywords: Bézier-like cubic function, Curvature continuity, Cshapedtransition curve, Spur gear tooth.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2295

Authors: Arturo Aquino, Manuel Emilio Gegundez, Diego Marin

Abstract:

In this paper, a new automated methodology to detect the optic disc (OD) automatically in retinal images from patients with risk of being affected by Diabetic Retinopathy (DR) and Macular Edema (ME) is presented. The detection procedure comprises two independent methodologies. On one hand, a location methodology obtains a pixel that belongs to the OD using image contrast analysis and structure filtering techniques and, on the other hand, a boundary segmentation methodology estimates a circular approximation of the OD boundary by applying mathematical morphology, edge detection techniques and the Circular Hough Transform. The methodologies were tested on a set of 1200 images composed of 229 retinographies from patients affected by DR with risk of ME, 431 with DR and no risk of ME and 540 images of healthy retinas. The location methodology obtained 98.83% success rate, whereas the OD boundary segmentation methodology obtained good circular OD boundary approximation in 94.58% of cases. The average computational time measured over the total set was 1.67 seconds for OD location and 5.78 seconds for OD boundary segmentation.

Keywords: Diabetic retinopathy, macular edema, optic disc, automated detection, automated segmentation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2742

Authors: V.K. Mukhomorov

Abstract:

The solvated electron is self-trapped (polaron) owing to strong interaction with the quantum polarization field. If the electron and quantum field are strongly coupled then the collective localized state of the field and quasi-particle is formed. In such a formation the electron motion is rather intricate. On the one hand the electron oscillated within a rather deep polarization potential well and undergoes the optical transitions, and on the other, it moves together with the center of inertia of the system and participates in the thermal random walk. The problem is to separate these motions correctly, rigorously taking into account the conservation laws. This can be conveniently done using Bogolyubov-Tyablikov method of canonical transformation to the collective coordinates. This transformation removes the translational degeneracy and allows one to develop the successive approximation algorithm for the energy and wave function while simultaneously fulfilling the law of conservation of total momentum of the system. The resulting equations determine the electron transitions and depend explicitly on the translational velocity of the quasi-particle as whole. The frequency of optical transition is calculated for the solvated electron in ammonia, and an estimate is made for the thermal-induced spectral bandwidth.

Keywords: Canonical transformations, solvated electron, width of the optical spectrum.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1264

Authors: Li Long, Thomas Ortlepp

Abstract:

A theoretical model for the optimization of thermopile sensor performance is developed for thermoelectric-based infrared radiation detection. It is shown that the performance of polycrystalline silicon film thermopile sensor can be optimized according to the thermoelectric quality factor, sensor layer structure factor and sensor layout shape factor. Based on the properties of electrons, phonons, grain boundaries and their interactions, the thermoelectric quality factor of polycrystalline silicon is analyzed with the relaxation time approximation of Boltzmann transport equation. The model includes the effects of grain structure, grain boundary trap properties and doping concentration. The layer structure factor of sensor is analyzed with respect to infrared absorption coefficient. The effect of layout design is characterized with the shape factor, which is calculated for different sensor designs. Double layer polycrystalline silicon thermopile infrared sensors on suspended support membrane have been designed and fabricated with a CMOS-compatible process. The theoretical approach is confirmed with measurement results.

Keywords: Polycrystalline silicon film, relaxation time approximation, specific detectivity, thermal conductivity, thermopile infrared sensor.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 150

Authors: Md. Sohidul Islam, Mohammad Rakibul Islam

Abstract:

In this paper, we consider a non-identically independently distributed (non-i.i.d.) Hoyt fading single-input multiple-out put (SIMO) channel, where the transmitter sends some confidential information to the legitimate receiver in presence of an eavesdropper. We formulated the probability of non-zero secrecy mutual information; secure outage probability and average secrecy mutual information (SMI) for the SIMO wireless communication system. The calculation has been carried out using small limit argument approximation (SLAA) on zeroth-order modified Bessel function of first kind. In our proposed model, an eavesdropper observes transmissions of information through another Hoyt fading channel. First, we derived the analytical expression for non-zero secrecy mutual information. Then, we find the secure outage probability to investigate the outage behavior of the proposed model. Finally, we find the average secrecy mutual information. We consider that the channel state information (CSI) is known to legitimate receiver.

Keywords: Hoyt fading, main channel, eavesdropper channel, secure outage probability, average secrecy mutual information.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1353

Authors: Zeitoun D. G., Thierry Dana-Picard

Abstract:

The study of a real function of two real variables can be supported by visualization using a Computer Algebra System (CAS). One type of constraints of the system is due to the algorithms implemented, yielding continuous approximations of the given function by interpolation. This often masks discontinuities of the function and can provide strange plots, not compatible with the mathematics. In recent years, point based geometry has gained increasing attention as an alternative surface representation, both for efficient rendering and for flexible geometry processing of complex surfaces. In this paper we present different artifacts created by mesh surfaces near discontinuities and propose a point based method that controls and reduces these artifacts. A least squares penalty method for an automatic generation of the mesh that controls the behavior of the chosen function is presented. The special feature of this method is the ability to improve the accuracy of the surface visualization near a set of interior points where the function may be discontinuous. The present method is formulated as a minimax problem and the non uniform mesh is generated using an iterative algorithm. Results show that for large poorly conditioned matrices, the new algorithm gives more accurate results than the classical preconditioned conjugate algorithm.

Keywords: Function singularities, mesh generation, point allocation, visualization, collocation least squares method, Augmented Lagrangian method, Uzawa's Algorithm, Preconditioned Conjugate Gradien

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1668

Authors: Myung-Gon Yoon, Jung-Ho Moon, Tae Kwon Ha

Abstract:

This paper studies a robust stabilization problem of a single agent in a multi-agent consensus system composed of identical agents, when the network topology of the system is completely unknown. It is shown that the transfer function of an agent in a consensus system can be described as a multiplicative perturbation of the isolated agent transfer function in frequency domain. From an existing robust stabilization result, we present sufficient conditions for a robust stabilization of an agent against unknown network topology.

Keywords: Multi-agent System, Robust Stabilization, Transfer Function.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1828

Authors: M. R. Aghaebrahimi, S. H. Zahiri, M. Amiri

Abstract:

The necessity of solving multi dimensional complicated scientific problems beside the necessity of several objective functions optimization are the most motive reason of born of artificial intelligence and heuristic methods. In this paper, we introduce a new method for multiobjective optimization based on learning automata. In the proposed method, search space divides into separate hyper-cubes and each cube is considered as an action. After gathering of all objective functions with separate weights, the cumulative function is considered as the fitness function. By the application of all the cubes to the cumulative function, we calculate the amount of amplification of each action and the algorithm continues its way to find the best solutions. In this Method, a lateral memory is used to gather the significant points of each iteration of the algorithm. Finally, by considering the domination factor, pareto front is estimated. Results of several experiments show the effectiveness of this method in comparison with genetic algorithm based method.

Keywords: Function optimization, Multiobjective optimization, Learning automata.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1638

Authors: Arkady Bolotin

Abstract:

One-way functions are functions that are easy to compute but hard to invert. Their existence is an open conjecture; it would imply the existence of intractable problems (i.e. NP-problems which are not in the P complexity class). If true, the existence of one-way functions would have an impact on the theoretical framework of physics, in particularly, quantum mechanics. Such aspect of one-way functions has never been shown before. In the present work, we put forward the following. We can calculate the microscopic state (say, the particle spin in the z direction) of a macroscopic system (a measuring apparatus registering the particle z-spin) by the system macroscopic state (the apparatus output); let us call this association the function F. The question is: can we compute the function F in the inverse direction? In other words, can we compute the macroscopic state of the system through its microscopic state (the preimage F -1)? In the paper, we assume that the function F is a one-way function. The assumption implies that at the macroscopic level the Schrödinger equation becomes unfeasible to compute. This unfeasibility plays a role of limit of the validity of the linear Schrödinger equation.

Keywords: One-way functions, P versus NP problem, quantummeasurements.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1265

Authors: A. Abedian, M. H. Ghiasi, B. Dehghan-Manshadi

Abstract:

A stiffened laminated composite panel (1 m length × 0.5m width) was optimized for minimum weight and deflection under several constraints using genetic algorithm. Here, a significant study on the performance of a penalty function with two kinds of static and dynamic penalty factors was conducted. The results have shown that linear dynamic penalty factors are more effective than the static ones. Also, a specially combined linear-exponential function has shown to perform more effective than the previously mentioned penalty functions. This was then resulted in the less sensitivity of the GA to the amount of penalty factor.

Keywords: Genetic algorithms, penalty function, stiffenedcomposite panel, finite element method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1630

Authors: Ephraim Gower, Thato Tsalaile, Monageng Kgwadi, Malcolm Hawksford.

Abstract:

In this paper, the statistical properties of filtered or convolved signals are considered by deriving the resulting density functions as well as the exact mean and variance expressions given a prior knowledge about the statistics of the individual signals in the filtering or convolution process. It is shown that the density function after linear convolution is a mixture density, where the number of density components is equal to the number of observations of the shortest signal. For circular convolution, the observed samples are characterized by a single density function, which is a sum of products.

Keywords: Circular Convolution, linear Convolution, mixture density function.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1472

Authors: Omid Mirmotahari, Yngvar Berg, Ahmad Habibizad Navin

Abstract:

Reversible logic is becoming more and more prominent as the technology sets higher demands on heat, power, scaling and stability. Reversible gates are able at any time to "undo" the current step or function. Multiple-valued logic has the advantage of transporting and evaluating higher bits each clock cycle than binary. Moreover, we demonstrate in this paper, combining these disciplines we can construct powerful multiple-valued reversible logic structures. In this paper a reversible block implemented by pseudo floatinggate can perform AD-function and a DA-function as its reverse application.

Keywords: Reversible logic, bi-directional, Pseudo floating-gate(PFG), multiple-valued logic (MVL).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1557

Authors: Nora Taghavi, Ali Sadr

Abstract:

System identification is the process of creating models of dynamic process from input- output signals. The aim of system identification can be identified as “ to find a model with adjustable parameters and then to adjust them so that the predicted output matches the measured output". This paper presents a method of modeling and simulating with system identification to achieve the maximum fitness for transformation function. First by using optimized KLM equivalent circuit for PVDF piezoelectric transducer and assuming different inputs including: sinuside, step and sum of sinusides, get the outputs, then by using system identification toolbox in MATLAB, we estimate the transformation function from inputs and outputs resulted in last program. Then compare the fitness of transformation function resulted from using ARX,OE(Output- Error) and BJ(Box-Jenkins) models in system identification toolbox and primary transformation function form KLM equivalent circuit.

Keywords: PVDF modeling, ARX, BJ(Box-Jenkins), OE(Output-Error), System Identification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2691

Authors: K. Soldatova, Y. Galerkin

Abstract:

A loading factor performance is necessary for the modeling of centrifugal compressor gas dynamic performance curve. Measured loading factors are linear function of a flow coefficient at an impeller exit. The performance does not depend on the compressibility criterion. To simulate loading factor performances, the authors present two parameters: a loading factor at zero flow rate and an angle between an ordinate and performance line. The calculated loading factor performances of non-viscous are linear too and close to experimental performances. Loading factor performances of several dozens of impellers with different blade exit angles, blade thickness and number, ratio of blade exit/inlet height, and two different type of blade mean line configuration. There are some trends of influence, which are evident – comparatively small blade thickness influence, and influence of geometry parameters is more for impellers with bigger blade exit angles, etc. Approximating equations for both parameters are suggested. The next phase of work will be simulating of experimental performances with the suggested approximation equations as a base.

Keywords: Centrifugal compressor stage, centrifugal compressor, loading factor, gas dynamic performance curve.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2047

Authors: Bhanu Gupta

Abstract:

In this paper, the criteria of Ψ-eventual stability have been established for generalized impulsive differential systems of multiple dependent variables. The sufficient conditions have been obtained using piecewise continuous Lyapunov function. An example is given to support our theoretical result.

Keywords: impulsive differential equations, Lyapunov function, eventual stability

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3992

Authors: Guangpu Chen

Abstract:

When the failure function is monotone, some monotonic reliability methods are used to gratefully simplify and facilitate the reliability computations. However, these methods often work in a transformed iso-probabilistic space. To this end, a monotonic simulator or transformation is needed in order that the transformed failure function is still monotone. This note proves at first that the output distribution of failure function is invariant under the transformation. And then it presents some conditions under which the transformed function is still monotone in the newly obtained space. These concern the copulas and the dependence concepts. In many engineering applications, the Gaussian copulas are often used to approximate the real word copulas while the available information on the random variables is limited to the set of marginal distributions and the covariances. So this note catches an importance on the conditional monotonicity of the often used transformation from an independent random vector into a dependent random vector with Gaussian copulas.

Keywords: Monotonic, Rosenblatt, Nataf transformation, dependence concepts, completely positive matrices, Gaussiancopulas

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1170

Authors: Edward Aris D. Fajardo, Hamdi Muhyuddin D. Barra

Abstract:

This study employs the use of the fourth order Numerov scheme to determine the eigenstates and eigenvalues of particles, electrons in particular, in single and double delta function potentials. For the single delta potential, it is found that the eigenstates could only be attained by using specific potential depths. The depth of the delta potential well has a value that varies depending on the delta strength. These depths are used for each well on the double delta function potential and the eigenvalues are determined. There are two bound states found in the computation, one with a symmetric eigenstate and another one which is antisymmetric.

Keywords: Double Delta Potential, Eigenstates, Eigenvalue, Numerov Method, Single Delta Potential

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3014

Authors: Wei Zhang, Yao-Yu Li, Yi-Fan Zhu, Qun Li, Wei-Ping Wang

Abstract:

It-s known that incorporating prior knowledge into support vector regression (SVR) can help to improve the approximation performance. Most of researches are concerned with the incorporation of knowledge in the form of numerical relationships. Little work, however, has been done to incorporate the prior knowledge on the structural relationships among the variables (referred as to Structural Prior Knowledge, SPK). This paper explores the incorporation of SPK in SVR by constructing appropriate admissible support vector kernel (SV kernel) based on the properties of reproducing kernel (R.K). Three-levels specifications of SPK are studied with the corresponding sub-levels of prior knowledge that can be considered for the method. These include Hierarchical SPK (HSPK), Interactional SPK (ISPK) consisting of independence, global and local interaction, Functional SPK (FSPK) composed of exterior-FSPK and interior-FSPK. A convenient tool for describing the SPK, namely Description Matrix of SPK is introduced. Subsequently, a new SVR, namely Motivated Support Vector Regression (MSVR) whose structure is motivated in part by SPK, is proposed. Synthetic examples show that it is possible to incorporate a wide variety of SPK and helpful to improve the approximation performance in complex cases. The benefits of MSVR are finally shown on a real-life military application, Air-toground battle simulation, which shows great potential for MSVR to the complex military applications.

Keywords: admissible support vector kernel, reproducing kernel, structural prior knowledge, motivated support vector regression

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1577

Authors: Wei Zhang, Yao-Yu Li, Yi-Fan Zhu, Qun Li, Wei-Ping Wang

Abstract:

It-s known that incorporating prior knowledge into support vector regression (SVR) can help to improve the approximation performance. Most of researches are concerned with the incorporation of knowledge in form of numerical relationships. Little work, however, has been done to incorporate the prior knowledge on the structural relationships among the variables (referred as to Structural Prior Knowledge, SPK). This paper explores the incorporation of SPK in SVR by constructing appropriate admissible support vector kernel (SV kernel) based on the properties of reproducing kernel (R.K). Three-levels specifications of SPK are studies with the corresponding sub-levels of prior knowledge that can be considered for the method. These include Hierarchical SPK (HSPK), Interactional SPK (ISPK) consisting of independence, global and local interaction, Functional SPK (FSPK) composed of exterior-FSPK and interior-FSPK. A convenient tool for describing the SPK, namely Description Matrix of SPK is introduced. Subsequently, a new SVR, namely Motivated Support Vector Regression (MSVR) whose structure is motivated in part by SPK, is proposed. Synthetic examples show that it is possible to incorporate a wide variety of SPK and helpful to improve the approximation performance in complex cases. The benefits of MSVR are finally shown on a real-life military application, Air-toground battle simulation, which shows great potential for MSVR to the complex military applications.

Keywords: admissible support vector kernel, reproducing kernel, structural prior knowledge, motivated support vector regression

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1353

Authors: Kourosh Parand, Zahra Delafkar, Fatemeh Baharifard

Abstract:

The problem of natural convection about a cone embedded in a porous medium at local Rayleigh numbers based on the boundary layer approximation and the Darcy-s law have been studied before. Similarity solutions for a full cone with the prescribed wall temperature or surface heat flux boundary conditions which is the power function of distance from the vertex of the inverted cone give us a third-order nonlinear differential equation. In this paper, an approximate method for solving higher-order ordinary differential equations is proposed. The approach is based on a rational Chebyshev Tau (RCT) method. The operational matrices of the derivative and product of rational Chebyshev (RC) functions are presented. These matrices together with the Tau method are utilized to reduce the solution of the higher-order ordinary differential equations to the solution of a system of algebraic equations. We also present the comparison of this work with others and show that the present method is applicable.

Keywords: Tau method, semi-infinite, nonlinear ODE, rational Chebyshev, porous media.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1884

Authors: Li Ni, Pen ManMan, Li KenLi

Abstract:

Clustering is an intensive research for some years because of its multifaceted applications, such as biology, information retrieval, medicine, business and so on. The expectation maximization (EM) is a kind of algorithm framework in clustering methods, one of the ten algorithms of machine learning. Traditionally, optimization of objective function has been the standard approach in EM. Hence, research has investigated the utility of evolutionary computing and related techniques in the regard. Chemical Reaction Optimization (CRO) is a recently established method. So the property embedded in CRO is used to solve optimization problems. This paper presents an algorithm framework (EM-CRO) with modified CRO operators based on EM cluster problems. The hybrid algorithm is mainly to solve the problem of initial value sensitivity of the objective function optimization clustering algorithm. Our experiments mainly take the EM classic algorithm:k-means and fuzzy k-means as an example, through the CRO algorithm to optimize its initial value, get K-means-CRO and FKM-CRO algorithm. The experimental results of them show that there is improved efficiency for solving objective function optimization clustering problems.

Keywords: Chemical reaction optimization, expectation maximization, initial, objective function clustering.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1235

Authors: Jasna Prester, Ivana Rašić Bakarić, Božidar Matijević

Abstract:

Innovations not only contribute to competitiveness of the company but have also positive effects on revenues. On average, product innovations account to 14 percent of companies’ sales. Innovation management has substantially changed during the last decade, because of growing reliance on external partners. As a consequence, a new task for purchasing arises, as firms need to understand which suppliers actually do have high potential contributing to the innovativeness of the firm and which do not. Proper organization of the purchasing function is important since for the majority of manufacturing companies deal with substantial material costs which pass through the purchasing function. In the past the purchasing function was largely seen as a transaction-oriented, clerical function but today purchasing is the intermediate with supply chain partners contributing to innovations, be it product or process innovations. Therefore, purchasing function has to be organized differently to enable firm innovation potential. However, innovations are inherently risky. There are behavioral risk (that some partner will take advantage of the other party), technological risk in terms of complexity of products and processes of manufacturing and incoming materials and finally market risks, which in fact judge the value of the innovation. These risks are investigated in this work. Specifically, technological risks which deal with complexity of the products, and processes will be investigated more thoroughly. Buying components or such high edge technologies necessities careful investigation of technical features and therefore is usually conducted by a team of experts. Therefore it is hypothesized that higher the technological risk, higher will be the centralization of the purchasing function as an interface with other supply chain members. Main contribution of this research lies is in the fact that analysis was performed on a large data set of 1493 companies, from 25 countries collected in the GMRG 4 survey. Most analyses of purchasing function are done by case study analysis of innovative firms. Therefore this study contributes with empirical evaluations that can be generalized.

Keywords: Purchasing function organization, innovation, technological risk, GMRG 4 survey.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3680

Authors: A. Selmi, A. Bisharat

Abstract:

Using Fourier transform and based on the Mindlin's 2^nd gradient model that involves two length scale parameters, the Green's function, the Eshelby tensor, and the Eshelby-like tensor for a spherical inclusion are derived. It is proved that the Eshelby tensor consists of two parts; the classical Eshelby tensor and a gradient part including the length scale parameters which enable the interpretation of the size effect. When the strain gradient is not taken into account, the obtained Green's function and Eshelby tensor reduce to its analogue based on the classical elasticity. The Eshelby tensor in and outside the inclusion, the volume average of the gradient part and the Eshelby-like tensor are explicitly obtained. Unlike the classical Eshelby tensor, the results show that the components of the new Eshelby tensor vary with the position and the inclusion dimensions. It is demonstrated that the contribution of the gradient part should not be neglected.

Keywords: Eshelby tensor, Eshelby-like tensor, Green’s function, Mindlin’s 2nd gradient model, Spherical inclusion.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 676