Search

Recently, transformer-based foundation models have proven to be a generalized architecture applicable to various data modalities, ranging from text to audio and even a combination of multiple modalities. Transformers by design should accurately model the non-trivial structure of particle showers thanks to the absence of strong inductive bias, better modeling of long-range dependencies, and interpolation and extrapolation capabilities. In this paper, we explore a transformer-based generative model for detector-agnostic fast shower simulation, where the goal is to generate synthetic particle showers, i.e., the energy depositions in the calorimeter. When trained with an adequate amount and variety of showers, these models should learn better representations compared to other deep learning models, and hence should quickly adapt to new detectors. In this work, we will show the prototype of a transformer-based generative model for fast shower simulation, as well as explore certain aspects of transformer architecture such as input data representation, sequence formation, and the learning mechanism for our unconventional shower data. [ABSTRACT FROM AUTHOR]

The paper considers a problem of extrapolating functions of several variables. It is assumed that the values of the function of m variables at a finite number of points in some domain D of the m-dimensional space are given. It is required to restore the value of the function at points outside the domain D. The paper proposes a fundamentally new method for functions of several variables extrapolation. In the presented paper, the method of extrapolating a function of many variables developed by us uses the interpolation scheme of metric analysis. To solve the extrapolation problem, a scheme based on metric analysis methods is proposed. This scheme consists of two stages. In the first stage, using the metric analysis, the function is interpolated to the points of the domain D belonging to the segment of the straight line connecting the center of the domain D with the point M, in which it is necessary to restore the value of the function. In the second stage, based on the auto regression model and metric analysis, the function values are predicted along the above straight-line segment beyond the domain D up to the point M. The presented numerical example demonstrates the efficiency of the method under consideration. [ABSTRACT FROM AUTHOR]

Autism is neurological disorder in which a person is affected with communication and interaction abilities. Lack of social interaction, repetitive behavior, and stable interest are indications of the autistic child. It is essential to identify the autism at very early stage. Machine learning plays vital role in health care which requires a process that reduces cost and time. The main aim of this paper is to implement machine learning algorithms and predict the disorder of autism. In this study, the different classification methods for diagnosing ASD were used on children of age 4 to 11 years. The present study proposes Support Vector Machine, K nearest neighbor, Decision tree and Linear Discriminant Analysis algorithm of machine learning to classify the autism spectrum disorder. Number of feature are extracted from the dataset using algorithm and statistically analyzed. Dataset was separated into 70:30 ratios. After applying above mentioned algorithm, comparison of various performance measures was done. It is observed that Decision trees and SVM gives higher accuracy of approx. 99% better than KNN and LDA with 70% and 97% respectively. Performance of classifiers is calculated using python which is generally used for machine learning classifications. [ABSTRACT FROM AUTHOR]

The paper presents a method for determining the force characteristics of the cutting process, introduced in the format of a modular system. The modules of input data generation, experiment planning, and results processing are highlighted and considered. The developed modular system makes it possible to obtain a polynomial dependence of the force characteristics on the processing parameters in a given technological area, which guarantees high accuracy of analytical calculations and eliminates errors associated with extrapolation. Within the modules, several features are considered, ensuring the connection of mathematical modeling and mechanical processing processes. In particular, the definition of the most significant parameters, the imposition of technological limitations, the use of multifactorial experiments, including fractional factorial ones, methods of decoding dependencies into natural values, and errors that arise in this case are considered. The use of the proposed modular system is demonstrated by examples of external longitudinal turning, drilling, and threading with taps and threaded cutters. It is shown that the use of this modular system makes it possible to predict the values of the cutting force within the factor space with an error of 2%. [ABSTRACT FROM AUTHOR]

The time-fractional Black-Scholes (TFBS) model is often used in the modeling of the fractional structure observed in financial markets. In this paper we propose a high-order compact numerical method for the solution of the pricing equation of the TFBS model. The proposed method uses a fourth-order compact three-point approximation in space and an implicit Euler discretization in time combined with an extrapolation technique. Numerical examples are carried out to show the merits of the method for pricing European and knock-out options under the TFBS model arising in the financial market following the discovery of fractals. [ABSTRACT FROM AUTHOR]

In different climatic and hydrological regional studies, data from grid databases that cover the climatic data of larger areas can be used. The authors of this contribution are interested in the region of Central Europe, where they can use the publicly available CarpatClim and ECA&D databases for such needs. The first mentioned database provides better spatial detail of the data, but the data is available only to the year 2010. On the other hand, the EC&D database is continually updated (up to the present), but the spatial detail in which it provides data is lower. This paper discusses the possibility of extending the time range of the CarpatClim climate data with help of the ECA&D database to the present. The data conversion was done using regression methods that will be described in more detail in the article. The authors also tested the usability of the Random Forest algorithm for this purpose. The quality of such data extrapolation was evaluated by the Pearson correlation coefficient R, which for the most problematic and consequently the least correlated climatic data, i.e., precipitation, has an average value of 0.91. For the best-correlated data, i.e., temperatures, the degree of accuracy determined by R is higher than 0.95. [ABSTRACT FROM AUTHOR]

This paper deals with the power production from a piezoelectric biodynamic harvesting scheme via the local step optimization point of the algorithm. In addition, the sensitivity and error of the maximum power point tracking (MPPT) algorithm are determined in this work due to its non-linear extrapolation approach in order to provide a real-time adaptation of the circuit impulses and to obtain matching of the target load impedance. The presented method provides a non-linear extrapolation model for the combined harvester design based on an artificial neural network (ANN). An analytical model of harvestable electrical power for piezoelectric vibration energy generation is proposed. Furthermore, steering stability is enhanced by a strong active front steering and an active differential drive control. The applicability of this approach outperformed the conventional method of observing perturbations under various operational requirements. [ABSTRACT FROM AUTHOR]

In video communications, the compressed video stream is very likely to be corrupted by channel errors. Recently, many error concealment algorithms have been proposed in order to combat channel errors. In this paper, we have combined two state-of-the-art motion vector recovery algorithms into a even better algorithm; we have modified the hybrid motion vector extrapolation (HMVE) algorithm and further improved the accuracy of the derived motion vectors by pixel-based MV correction using neighboring received MVs. Experimental results show that this method is able to provide better performance than existing algorithms subjectively and objectively. [ABSTRACT FROM PUBLISHER]

Estimating remaining useful life (RUL) is a crucial part in a successful online monitoring system. Extrapolation of a degradation model based on particle filtering (PF) approach, which is implemented on state-space model, is a popular method to predict RUL. Taking into account the characteristics of state-space model, the initial setting of process and measurement noise has a great impact on the predicted result. This paper discusses the individual performance of two popular PFs, sequential importance resampling PF and auxiliary PF, when they come to different noise characteristics. Two groups of initial process and measurement noises were set to compare the predicted performance between these two PFs. The prediction of battery RUL was demonstrated in this paper as a case study. The comparative results were used for reference to other-related degradation components or system using PFs-based prediction. [ABSTRACT FROM PUBLISHER]

Many prognosis studies have been conducted for a long time. There are many established and widely accepted prediction methods, such as linear extrapolation and SARIMA. However, their performance is far from perfect, especially when the time series is highly volatile. In this paper, we propose a hybrid prediction scheme that combines the classical SARIMA method and the wavelet transform (WT). Wavelet transform (WT) has emerged as an effective tool in decomposing time series into different components, which allows for improved prediction accuracy. However, this issue has so far been insufficiently tested and tried to predict different time series. Our goal is therefore to integrate modeling approaches as a decision support tool. The results of an empirical study show that this method can achieve high accuracy in prediction. Based on the results of the created model, it can be stated that the hybrid WSARIMA model overperformed the SARIMA model. [ABSTRACT FROM AUTHOR]

Lattice QCD calculations provide an ab initio access to hadronic process. These calculations are usu ally performed in a small cubic volume with periodic boundary conditions. The infinite volume extrapolations for three-body systems are indispensable to understand many systems of high current interest. We derive the three-body quantization condition in a finite volume using an effective field theory in the particle-dimer picture. Our work shows a powerful and transparent method to read off three-body physical observables from lattice simulations. In this paper, we review the formalism, quantization condition, spectrum analysis and energy shifts calculation both for 3-body bound states and scattering states. [ABSTRACT FROM AUTHOR]

In this paper we present a novel splitting approach to solve scale-dependent Schrödinger equations, which are applied in quantum mechanical models. The energy potential is based on highly oscillating functions and we obtain a multiscale problem. We apply a multiscale solver, which is based on an operator splitting method and decouples the diffusion and reaction part of the Schödinger equation. We apply a large time step for the implicit time-discretization of the diffusion part and small time steps for the explicit time-discretization of the reaction part. With an extrapolation step, we can reduce the computational time in the highly-oscillating time-scale. We present the numerical analysis of the extrapolated operator splitting method. First numerical experiments indicate the benefit of the extrapolated splitting approaches. [ABSTRACT FROM AUTHOR]

Extrapolation involves taking a certain linear combination of the numerical solutions of a base method applied with different stepsizes to obtain greater accuracy. This linear combination is done so as to eliminate the leading error term. The technique of extrapolation in accelerating convergence has been successfully in numerical solution of ordinary differential equations. In this study, symmetric Runge-Kutta methods for solving linear and nonlinear stiff problem are considered. Symmetric methods admit asymptotic error expansion in even powers of the stepsize and are therefore of special interest because successive extrapolations can increase the order by two at time. Although extrapolation can give greater accuracy, due to the stepsize chosen, the numerical approximations are often destroy due to the accumulated round off errors. Therefore, it is important to control the rounding errors especially when applying extrapolation. One way to minimize round off errors is by applying compensated summation. In this paper, the numerical results are given for the symmetric Runge-Kutta methods Implicit Midpoint and Implicit Trapezoidal Rule applied with and without compensated summation. The result shows that symmetric methods with higher level extrapolation using compensated summation gives much smaller errors. On the other hand, symmetric methods without compensated summation when applied with extrapolation, the errors are affected badly by rounding errors. [ABSTRACT FROM AUTHOR]

The 13C(α,n)16O reaction is the prevalent neutron source for the main s-process. The direct measurement of this reaction at stellar temperature (kT=8 keV) has so far not been possible due to the very low cross section at the corresponding energy. The extrapolation of the astrophysical S-factor of this reaction into the Gamow window (Eα,c.m.=140-230 keV) is complicated by the large uncertainties of the low-energy experimental data and the existence of a state of 17O near the α-threshold that can have a large effect on low energy cross section. The aim of this paper is to introduce the new LUNA experimental setup, dedicated to the investigation of 13C(α,n)16O reaction below Eα,lab=400 keV. [ABSTRACT FROM AUTHOR]

An improved classifier based on nearest feature line, is proposed in this paper. The new classifier is called as limited nearest feature line (LNFL). The purpose of LNFL is to improve the miss-classification of nearest feature line when the prototypes in NFL are far away from the query sample. A lot of experiments are executed on ORL and AR face database. And the detailed comparison result is given to show that LNFL is better than nearest feature line and nearest neighbor. [ABSTRACT FROM PUBLISHER]

Variations of the Channel State Information (CSI) between 2 CSI samples is a major source of performance degradation for Adaptive Modulation and Coding (AMC) schemes on fast fading channel. In this paper, we present a novel AMC design that highly improves the channel availability in those conditions. Based on a low complexity design which uses only one outdated CSI value, and without explicit knowledge of the fading process properties (rate and auto-correlation function), the proposed scheme significantly increases channel availability with low CSI update rate. [ABSTRACT FROM PUBLISHER]

In this paper, we address the problem of image bit-depth expansion and present a novel method to generate high bit-depth (HBD) images from a single low bit-depth (LBD) image. We expand image bit-depth by reconstructing the least significant bits (LSBs) for the LBD image after it is rescaled to high bit-depth. For image regions whose intensities are neither locally maximum nor minimum, neighborhood flooding is applied to convert 2D interpolation problem into 1D interpolation, for local maxima/minima (LMM) regions where interpolation is not applicable, a virtual skeleton marking algorithm is proposed to convert problematic 2D extrapolation problem into 1D interpolation. At last, a content-adaptive reconstruction model is proposed to obtain the output HBD image. The experimental results show that proposed method significantly outperforms existing methods in PSNR and SSIM without contouring artifacts. [ABSTRACT FROM PUBLISHER]

Nondestructive evaluation is used widely in many engineering and industrial areas to detect defects or flaws such as cracks inside parts or structures during manufacturing or for products in service. The standard statistical model is a simple empirical linear regression between the (possibly transformed) signal response variables and the (possibly transformed) explanatory variables. For some applications, such a simple empirical approach is inadequate. An important alternative approach is to use knowledge of the physics of the inspection process to provide information about the underlying relationship between the response and explanatory variables. Use of such knowledge can greatly increase the power and accuracy of the statistical analysis and enable, when needed, proper extrapolation outside the range of the observed explanatory variables. This paper describes a set of physical model-assisted analyses to study the capability of two different ultrasonic testing inspection methods to detect synthetic hard alpha inclusion and flat-bottom hole defects in a titanium forging disk. [ABSTRACT FROM AUTHOR]

This paper proposes novel lattice algorithms to compute tail conditional expectation. To improve the naive approach, we develop tilting, trinomial, and extrapolation algorithms together with some syntheses of these algorithms. Furthermore, we introduce fractional-step lattices to prevent interpolation error in the extrapolation algorithms. A key finding is that combining the techniques of tilting lattice, extrapolation and fractional steps substantially increases speed and accuracy. We demostrate the efficiency and accuracy of these algorithms with numerical results. [ABSTRACT FROM AUTHOR]

In this paper we present a statistical method to build three-dimensional gas distribution maps (3D-DM). The proposed mapping technique uses kernel extrapolation with a tri-variate Gaussian kernel that models the likelihood that a reading represents the concentration distribution at a distant location in the three dimensions. The method is evaluated using a mobile robot equipped with three “e-noses” mounted at different heights. Initial experiments in an uncontrolled indoor environment are presented and evaluated with respect to the ability of the 3D map, computed from the lower and upper nose, to predict the map from the middle nose. [ABSTRACT FROM AUTHOR]

The present paper is devoted to the numerical solution of the Black-Scholes equation for pricing European options. We apply the Crank-Nicolson scheme with Richardson extrapolation for time discretization and Hermite cubic spline wavelets with four vanishing moments for space discretization. This scheme is the fourth order accurate both in time and in space. Computational results indicate that the Crank-Nicolson scheme with Richardson extrapolation significantly decreases the amount of computational work. We also numerically show that optimal convergence rate for the used scheme is obtained without using startup procedure despite the data irregularities in the model. [ABSTRACT FROM AUTHOR]

Short time series prediction technique based on Bernstein polynomials is presented in this paper. Firstly, the straightforward Bernstein polynomial extrapolation scheme is improved by extending the interval of approximation. Secondly, the forecasting scheme is designed in the evolutionary computational setup which is based on the conciliation between the coarseness of the algebraic prediction and the smoothness of the time average prediction. Computational experiments with the test time series suggest that this time series prediction technique could be applicable for various forecasting applications. [ABSTRACT FROM AUTHOR]

Proper spatial sampling is critical for many applications. If the sampling criterion is not met, artifacts appear for example in images. Last year an iterative approach was presented using wave field extrapolation to interpolate spatially aliased signals. The main idea behind this approach is that after inverse wave field extrapolation the signal is concentrated in a small region with a high amplitude, while the aliasing artifacts are spread-out through the domain. Inverse wave field extrapolation focusses optimally at one depth, making the performance of the reconstruction depth dependent. Obviously the method can be repeated for several depths. This year we show an alternative approach using an imaging/inverse-imaging approach. The demonstration of this approach is extended to 2D-arrays where the sampling limitations are even more critical. Moreover we show in this paper that the interpolation approach is not limited to near-field data, but can also be used on far-field data (plane waves). The Radon transform can be used for plane waves to focus the data. The approach is demonstrated using modeled and measured data for linear and 2D arrays. [ABSTRACT FROM AUTHOR]

We propose sixth-order approximations for semilinear parabolic systems that are based on fourth-order compact difference schemes combined with Richardson extrapolation. Within this paper, Richardson extrapolation is applied directly to the solution procedures used within existing numerical algorithms for realization of compact fourth-order difference schemes. Numerical results are presented to demonstrate the accuracy and efficiency of the proposed finite difference discretization strategy. [ABSTRACT FROM AUTHOR]

The zero-shear-rate viscosity η0 is a limiting value that cannot be measured directly; rather, it must be estimated by extrapolation. The value determined by extrapolation will depend on the model used, and the range and quality of the data used for the extrapolation. The error can be surprisingly large. The ongoing study described in this paper has several objectives including development of methods for bracketing the true value, and for improving the estimation process with less-than-ideal sets of data. These methods have been evaluated using data for two room-temperature polymer melts: poly(dimethyl siloxane) (PDMS) and poly(isobutylene) (PIB). [ABSTRACT FROM AUTHOR]

We consider here a natural spline interpolation based on reduced data and the so-called exponential parameterization (depending on parameter α ϵ [0;1]). In particular, the latter is studied in the context of the trajectory approximation in arbitrary euclidean space. The term reduced data refers to an ordered collection of interpolation points without provision of the corresponding knots. The numerical verification of the intrinsic asymptotics α(ϵ) in γ approximation by natural spline ĝN3 is conducted here for regular and sufficiently smooth curve γ sampled more-or-less uniformly. We select in this paper the substitutes for the missing knots according to the exponential parameterization. The outcomes of the numerical tests manifest sharp linear convergence orders α(ϵ) = 1, for all α ϵ [0;1). In addition, the latter results in unexpected left-hand side discontinuity at α = 1, since as shown again here a sharp quadratic order α(1) = 2 prevails. Remarkably, the case of α(1) = 2 (derived for reduced data) coincides with the well-known asymptotics established for a natural spline to fit non-reduced data determined by the sequence of interpolation points supplemented with the corresponding knots (see e.g. [1]). [ABSTRACT FROM AUTHOR]

Richardson extrapolation is proposed about a century ago and broadly accepted in the scientific community as a mean to accelerate the convergence of approximate computations. Many implementations, generalizations, and applications, of Richardson extrapolation, are reported in the literature. In this paper, considering the Taylor series expansion of approximate computations with respect to the algorithmic parameters, Richardson extrapolation is introduced as a tool towards higher accuracies, and the higher orders of accuracy are achieved as by-products. On this basis, a new proposition on Richardson extrapolation is stated, explained, and demonstrated, in an example. [ABSTRACT FROM AUTHOR]

In the Nepalese Himalayas, the complex topography, occurrence of debris covered glaciers, and limited data availability creates substantial difficulties for modelling glacier melt. The proper recognition of melt processes governs the accurate estimation of melt water from glacier dominated systems, even in the presence of debris-covered glaciers. This paper presents a glacier melt model developed for the Lirung sub-basin of Langtang valley, which has both a clean glacier area, 5.86 km², and a debris-covered glacier area, 1.13 km². We use a temperature index approach to estimate sub-daily melt water discharge for a two week period at the end of monsoon, and the melt factor is varied according to the aspect and distributed to each grid processed from the digital elevation model. The model uses easily available data and simple extrapolation techniques capable of generating melt with limited data. The result obtained from this method provides accurate estimate with an R² value of 0.89, bias of 0.9% and Nash-Sutcliffe efficiency of 0.86, and suitable in Himalaya where data availability is major issue. [ABSTRACT FROM AUTHOR]

In this paper, we have improved the limitations of our previous Richardson's extrapolation spreadsheet calculator for computing differentiations numerically. These limitations are the value of D(0,0) keyed in by users using 3-point central difference formula, and the fact that the previous spreadsheet calculator can only calculate the approximate definite differentiation up to level 4 x 4. If the function to be differentiated is complicated, calculating D(0,0) using 3-point central difference formula can be tedious as parentheses should be put in a proper order when writing the calculation command. Otherwise, the calculation command may lead to a wrong answer. In this improved Richardson's extrapolation spreadsheet calculator, we redesigned the Richardson's extrapolation spreadsheet calculator, where users are only required to give the value of x, the function to be differentiated f(x), and the step size h value without writing the command to obtain D(0,0). Consequently, the calculations will be done automatically to approximate the definite differentiation up to level 10 x 10. [ABSTRACT FROM AUTHOR]

Animating a character's face is a strenuous task due to a lack of a consensus among 3D artists on how to build and provide the controls in an animation-ready facial rig. To help provide a uniform integrative of control interaction to facial animation, a multimodal rig interface system approach was developed and tested in an initial user experiment with 15 digital artists. The users requested improvements and the system was upgraded into a user-adapted multimodal approach that supports a synchronized input of its different interaction modes and a scalable control of the facial deformation. This paper describes the algorithms behind the user-adapted system and a follow-up user experiment with 20 professional expert character animators. The results and user scores help clarify the relevance of the user-adapted multimodal rig interface system approach as an interaction model that decreases the effort required to animate a character's face. [ABSTRACT FROM AUTHOR]

Designers of adaptive systems for regression often have different kinds of a priori knowledge about the function that is to be learned. By using that knowledge, the learning process can be led to faster and better convergence and increased robustness. Those designers deal with the problem of how to effectively use that knowledge in the formulation of the learning problem. This paper deals with the formalization of using different kinds of local a priori knowledge for a specific form of local regularization of the learning problem. The GIESELA approach presented here enables the designer to utilize his local absolute, relative and qualitative a priori knowledge about the function to be learned in a robust and interpretable way and to apply this knowledge on a global scale. The approach is mathematically analyzed and demonstrated on a simple example. [ABSTRACT FROM AUTHOR]

Phased array inspection was conducted on a V-butt welded steel sample with multiple shallow flaws of varying depths. The inspection measurements were processed using Wiener filtering and Autoregressive Spectral Extrapolation (AS) to enhance the signals. Phased array inspections were conducted using multiple phased array probes of varying nominal central frequencies (2.25, 4, 5 and 10 MHz). This paper describes the measured results, which show high accuracy, typically in the range of 0.1-0.2 mm. The results concluded that: 1. There was no statistical difference between the calculated flaw depths from phased array inspections at different flaw tip angles. 2. There was no statistical difference in flaw depths calculated using phased array data collected from either side of the weld. 3. Flaws with depths less than the estimated probe signal shear wavelength could not be sized. 4. Finally, there was no statistical difference in the calculated flaw depths using phased array probes with different sampling frequencies and destructive measurements of the flaws. [ABSTRACT FROM AUTHOR]

The carbon fusion project at Notre Dame is aimed towards measuring the 12C+12C fusion cross section and its decay branches relevant to astrophysics down to the lowest possible energies. To complement this approach, we are also exploring new techniques for providing more reliable extrapolations of the cross sections in the energy ranges where experimental data are unavailable. In this paper, we report two recent results: 1) an upper limit for the 12C+12C fusion cross section, and 2) a new measurement of 12C(12C,n) along with an improved extrapolation technique based on the mirror reaction channel, 12C(12C,p). The outlook for astrophysical heavy-ion fusion studies at Notre Dame is also discussed. [ABSTRACT FROM AUTHOR]

In this paper, a novel classifier based nearest feature line, called as restricted nearest feature line with angle, is proposed. NFL was originally used in face recognition, and later began to be used in many other applications. Though successful in improving the classification ability, there are still some drawbacks in NFL that limit their further application in practice. To improve the miss-classification of nearest feature line when the prototypes in NFL are far away from the query sample, restricted nearest feature line with angle uses the angle between two lines to restrict the feature lines. The two lines are composed by query sample and both prototype samples. A large number of experiments are executed on ORL and AR face database. And detailed comparison result is given. [ABSTRACT FROM PUBLISHER]

The FORMOSAT-3/COSMIC satellite constellation has become an important source of remote sensing data globally distributed for the sounding of the atmosphere of the Earth and, in particular, the ionosphere. In this study, electron density profiles derived using the Improved Abel transform inversion in Radio Occultation (RO) scenarios are used as input data to derive some features regarding the topside and outside ionospheric contribution, hence, the plasmasphere in great extend, by means of the analysis of the integral values of the shape functions corresponding to each density profile. The novelty presented in this paper, with respect to previous works, is the use of experimental data from the FORMOSAT-3/COSMIC RO-derived electron densities to infer global characterizations and distribution of the Total Electron Content (TEC) into its main components: ionospheric TEC and plasmaspheric TEC, both contributing to TEC. The results show agreement with earlier modeled and observational data from previous local studies of the plasmaspheric contribution. The main conclusion of this research is that the plasmasphere contributes significantly to TEC and the ratio between plasmaspheric TEC and TEC has been climatologically analyzed for the whole year 2007 confirming that the major relative impact of the plasmasphere is during night time. The added value of this study is that the results obtained are globally distributed and can help to infer a proxy for the plasmaspheric contribution at any location over the globe thanks to the FORMOSAT-3/COSMIC coverage. [ABSTRACT FROM PUBLISHER]

In this paper, a novel manifold learning algorithm termed orthogonal nearest neighbor feature space embedding (ONNFSE) is proposed to eliminate three drawbacks of the nearest feature space embedding (NFSE) approach. The first one is an extrapolation error, a feature line passes through two far neighbor points is selected for scatter matrix calculating when the distance of a specified point to this line is small. The calculated scatter matrix could not efficiently preserve the local topological structure among samples. The incorrect selection will reduce the recognition rates. The interpolation error is similar the extrapolation one. To remedy these two problems, the nearest neighbor feature space is built in the proposed ONNFSE. The last problem should be solved is the non-orthogonal eigenvectors found by the NFSE algorithm. The modified ONNFSE algorithm generates orthogonal bases which possess the more discriminating power. Experimental results are conducted to demonstrate the effectiveness of our proposed algorithm. [ABSTRACT FROM PUBLISHER]

While the generalizability of classifiers receive much attention in research, interpretability is often neglected. This paper proposes a rule-plus-exemplar classification framework based on ideas in cognitive psychology. The classification process is interpretable and intuitive, and also generalizes well. It can perform better than other interpretable methods such as decision trees, for both interpolative and extrapolative generalization. [ABSTRACT FROM PUBLISHER]

This paper presents a case study of the interaction between nanoelectrodes and dielectric nanoparticles possibly representative of biomolecules in a simple cylindrical capacitive biosensor. The small signal admittance change due to the insertion of the biomolecule in the biosensor electrolyte is studied as a function of the position, aspect ratio and charge of the biomolecule and of the signal frequency. Results suggest clear advantages in operating the biosensor beyond the electrolyte cutoff frequency. [ABSTRACT FROM PUBLISHER]

Proper evaluation of load is the most important step for the design of a structure. This paper presents the uncertainty modeling of environmental load for jacket platforms in Malaysia. We are interested only in extreme loads for the modeling of environmental load which are distributed by extreme values. Here in this paper, Weibull or Gumbel distributions are used. The existing platforms are designed for 100 year return period but are assessed for 10,000 years as recommended by ISO 19902 and API RP2A. The nature of environmental load is probabilistic and we need to ascertain the randomness of the load. This study covers three parameters of environmental load acting on jacket platforms namely significant wave height (Hs), wind speed and surface current speed in the Peninsular Malaysia (PM) region. Two-parameter Weibull and Gumbel distributions were used for the analysis. The data is in shape of 1, 10, 50, 100 year return period. Parameters of distribution are determined first, which are used to find mean and standard deviation of respective distribution. From this, uncertainty variables are extrapolated up to 1000 and 10,000 years. Finally the data was compared with data from Gulf of Mexico (GOM), Northern North Sea (NNS), Southern North Sea (SNS) and Central North Sea (CNS). This work will lead to determining the reliability of jacket platforms in Malaysia. [ABSTRACT FROM PUBLISHER]

Recently, many types of cameras have come into use around the world. Because of this, color correction has become a significant problem and color charts have been developed for evaluation and correction purposes. In this paper, a method for detecting color charts of arbitrary size, rotation and location from camera images is proposed for color correction purposes. This method combines several hierarchical approaches with a straightforward algorithm to generate fast solutions. The effectiveness of this algorithm is confirmed using a 167-image dataset that includes several sizes and rotations of color chart placed at arbitrary locations. [ABSTRACT FROM PUBLISHER]

In this paper, ACM finite element method is applied to the approximation of nonlinear viscoelasticity equation for the semi-discrete scheme. by the integral identity, averaging technique and interpolated post processing techniques, the superclose and the global superconvergence properties with order O(h3.5) are derived. Furthermore, through constructing a suitable extrapolation scheme, higher accuracy of extrapolation solution is obtained. The results proposed in the paper are the same as that of the linear case in the previous literature. [ABSTRACT FROM PUBLISHER]

In 3D video coding, the Multi-View Video plus Depth (MVD) representation format enables the interpolation of virtual intermediate views in between the original recorded camera views, as well as the extrapolation of views outside this range using view synthesis algorithms. For this, MVD includes depth maps as additional information to be transmitted. The performance of the associated depth data coding can be improved by considering the distortion of synthesized views in the rate-distortion optimization of the encoder. Therefore, this paper evaluates coding gains achieved with encoder optimization with respect to virtual view positions. The impact of the number and the positions of virtual views used as reference in the encoder side optimization process on the quality of the synthesized views is analyzed. Coding results for view interpolation and view extrapolation are provided. Moreover, the encoder-side complexity with respect to the number of reference views is discussed. [ABSTRACT FROM PUBLISHER]

Fuzzy rule interpolation is well known for reducing the complexity of fuzzy models and making inference possible in sparse rule-based systems. However, in practical applications with inter-connected subsets of rules, situations may arise when a crucial antecedent of observation is absent, either due to human error or difficulty in obtaining data, while the associated conclusion may be derived according to alternative rules or even observed directly. If such missing antecedents were involved in the subsequent interpolation process, the final conclusion would not be deduced using conventional means. However, missing antecedents may be related to certain conclusion and therefore, may be inferred or interpolated using the known antecedents and conclusion. For this purpose, this paper presents a novel approach termed backward fuzzy rule interpolation and extrapolation. In particular, the approach supports both interpolation and extrapolation which involve multiple intertwined fuzzy rules, with each having multiple antecedents. An algorithm is given to implement the approach via the use of the scale and move transformation-based fuzzy interpolation. The algorithm makes use of trapezoidal fuzzy membership functions. Realistic application examples are provided to demonstrate the efficacy of the approach. [ABSTRACT FROM PUBLISHER]

Image alignment and mosaicing are usually performed on a set of overlapping images, using features in the area of overlap for alignment and for seamless stitching. Without image overlap current methods are helpless, and this is the case we address in this paper. So if a traveler wants to create a panoramic mosaic of a scene from pictures he has taken, but realizes back home that his pictures do not overlap, there is still hope. The proposed process has three stages: (i) Images are extrapolated beyond their original boundaries, hoping that the extrapolated areas will cover the gaps between them. This extrapolation becomes more blurred as we move away from the original image. (ii) The extrapolated images are aligned and their relative positions recovered. (iii) The gaps between the images are inpainted to create a seamless mosaic image. [ABSTRACT FROM PUBLISHER]

In pattern classification, the test data is expected to lie in the domain covered by the training data. But in practical scenarios, this may not necessarily be true. To improve the adaptability, the classifier should be able to generalize well even when there are changes in the input distribution. This paper proposes a cognitively inspired classification framework based on rules and exemplars. It can generalize well even for samples falling outside the region covered by the training data. [ABSTRACT FROM PUBLISHER]

Modified embedding interpolation and extrapolation model in Lanczos-types algorithms (MEIEMLA) is well-known as a strategy to avoid breakdown in Lanczos-type algorithms by taking advantage of the patterns which persistently exist in the iterates generated by a Lanczos-type algorithm. It is the improved EIEMLA by interpolating all vector solutions S = x1, x2,…,xk generated by a Lanczos-type algorithm, at one time by using PCHIP. Note, the previous EIEMLA interpolates the vector solutions by arranging S such that it contains all of the ith entries of iterates xk, for i = 1,2,…,n, and n is the dimension of the system. As a result, MEIEMLA performed the solutions of several systems of linear equations (SLEs) more efficient than EIEMLA. However, once running the MEIEMLA, it cannot convergent quickly. It is, therefore, important to put MEIEMLA in a restarting framework. In this paper, we combined the MEIEMLA and restarting in order for Lanczos-types algorithms getting rid off breakdown as well as enhancing their stability. Numerical results are presented and compared with other existing restarting strategies in Lanczos algorithms, including with REIEMLA. The results showed that the restarting MEIEMLA (RMEIEMLA) performed better than other existing restarting strategies. [ABSTRACT FROM AUTHOR]

In this paper, a new identification method for piecewise affine (PWA) models is introduced. The method is based on the data-based representation of PWA maps and the data compression with l1 optimization technique, which enable the method to deal with large data sets. This method can be applied to a wide range of modeling problems, and an example with a DC motor system is shown in this paper to show the usability of the method. [ABSTRACT FROM PUBLISHER]