Search results

The immense quantity of available unstructured text documents serve as one of the largest source of information. Text classification can be an essential task for many purposes in information retrieval, such as document organization, text filtering and sentiment analysis. Ensemble learning has been extensively studied to construct efficient text classification schemes with higher predictive performance and generalization ability. The purpose of this paper is to provide diversity among the classification algorithms of ensemble, which is a key issue in the ensemble design.

An ensemble scheme based on hybrid supervised clustering is presented for text classification. In the presented scheme, supervised hybrid clustering, which is based on cuckoo search algorithm and k-means, is introduced to partition the data samples of each class into clusters so that training subsets with higher diversities can be provided. Each classifier is trained on the diversified training subsets and the predictions of individual classifiers are combined by the majority voting rule. The predictive performance of the proposed classifier ensemble is compared to conventional classification algorithms (such as Naïve Bayes, logistic regression, support vector machines and C4.5 algorithm) and ensemble learning methods (such as AdaBoost, bagging and random subspace) using 11 text benchmarks.

The experimental results indicate that the presented classifier ensemble outperforms the conventional classification algorithms and ensemble learning methods for text classification.

The presented ensemble scheme is the first to use supervised clustering to obtain diverse ensemble for text classification

The purpose of this study is to construct and analyze a parameter uniform higher-order scheme for singularly perturbed delay parabolic problem (SPDPP) of convection-diffusion type with a multiple interior turning point.

The authors construct a higher-order numerical method comprised of a hybrid scheme on a generalized Shishkin mesh in space variable and the implicit Euler method on a uniform mesh in the time variable. The hybrid scheme is a combination of simple upwind scheme and the central difference scheme.

The proposed method has a convergence rate of order O(N−2L2+Δt). Further, Richardson extrapolation is used to obtain convergence rate of order two in the time variable. The hybrid scheme accompanied with extrapolation is second-order convergent in time and almost second-order convergent in space up to a logarithmic factor.

A class of SPDPPs of convection-diffusion type with a multiple interior turning point is studied in this paper. The exact solution of the considered class of problems exhibit two exponential boundary layers. The theoretical results are supported via conducting numerical experiments. The results obtained using the proposed scheme are also compared with the simple upwind scheme.

A mathematical model to predict the concentration polarisation in nanofiltration/reverse osmosis is described. It incorporates physical modelling for mass transfer, laminar hydrodynamics and the membrane rejection coefficient. The SIMPLE algorithm solves the discretised equations derived from the governing differential equations. The convection and diffusive terms of those equations are discretised by the upwind, the hybrid and the exponential schemes for comparison purposes. The hybrid scheme appears as the most suitable one for the type of flows studied herein. The model is first applied to predict the concentration polarisation in a slit, for which mathematical solutions for velocities and concentrations exist. Different grids are used within the hybrid scheme to evaluate the model sensitivity to the grid refinement. The 55×25 grid results agree excellently for engineering purposes with the known solutions. The model, incorporating a variation law for the membrane intrinsic rejection coefficient, was also applied to the predictions of a laboratory slit where experiments are performed and reported, yielding excellent results when compared with the experiments.

This paper aims to tackle the problem of thermo‐solutal convection and macrosegregation during ingot solidification of metal alloys. Complex flow structures associated with the development of channels segregate and sharp gradients in the solutal field call for the implementation of accurate methods for numerical modeling of alloy solidification. In particular, the solute transport equation is convection dominated and requires special non‐oscillarity type high‐order schemes to handle the regions of channels segregates.

In the present study, a time‐splitting approach has been adopted to separately handle solute advection and diffusion. This splitting technique allows the application of accurate total variation dimensioning (TVD) schemes for solution of solute advection. Applications of second‐order Lax‐Wendroff TVD SUPERBEE and fifth‐order weighted essentially non‐oscillatory (WENO) schemes are described in the present article. Classical numerical solution of solute transport using hybrid and central‐difference schemes are also employed for the purpose of comparisons. Numerical simulations for solidification of Pb‐18%Sn in a two‐dimensional rectangular cavity have been carried out using different numerical schemes.

Numerical results show the difficulty of obtaining grid‐independent solutions with respect to local details in the region of channels. Grid convergence patterns and numerical uncertainty are found to be dependent on the applied scheme. In general, the first‐order hybrid scheme is diffusive and under predicts the formation of channels. The second‐order central‐difference scheme brings about oscillations with possible non‐physical extremes of solute composition in the region of channel segregates due to sharp gradients in the solutal field. The results obtained using TVD and WENO schemes contain no oscillations and show an excellent capture of channels formation and resolution of the interface between solute‐rich and depleted bands. Different stages of channels formation are followed by analyzing thermo‐solutal convection and macrosegregation at different times during solidification.

Accurate prediction of local variation in the solutal and flow fields in the channels regions requires grid refinement up to scales in the order of microscopic dendrite arm spacing. This imposes limitations in terms of large computational time and applicability of available macroscopic models based on classical volume‐averaging techniques.

The present study is very useful for numerical simulation of macrosegregation during ingot casting of metal alloys.

The paper provides the methodology and application of TVD schemes to predict channel segregates during columnar solidification of metal alloys. It also demonstrates the limitations of classical schemes for simulation of alloy solidification.

This study aims to understand how the recent COVID-19 pandemic impacted burnout levels among working mothers in leadership positions and how income and work schemes play an important role in their burnout.

Data were collected from 961 working mothers and fathers in leadership positions in Mexico under different work schemes during the COVID-19 lockdown. Snowball sampling was used in this study. The Maslach Burnout Inventory General Survey was distributed online, using the burnout scale, with income and work schemes as categorical variables.

Burnout levels among working mothers in leadership positions were higher than those among working fathers. The hybrid work scheme (i.e. working from home combined with working from office) lessens burnout in working mothers, contributing both theoretically and empirically to better understanding burnout levels of mothers in leadership positions.

The findings can encourage human resource areas to reflect on the overexertion and work stress of mothers in leadership positions, and potential support resources can be provided to motivate them and retain their talent.

The introduction of the notion of lockdown in a conceptual model to observe its interaction with burnout and hybrid work schemes (i.e. working from the office and home) has rarely been discussed in existing literature. The impact, especially for working mothers in leadership positions, must be thus carefully considered while dealing with future crises, thereby helping to develop policies and processes accordingly.

The study based on the estimation of the stress–strength reliability parameter plays a vital role in showing system efficiency. In this paper, considering independent strength and stress random variables distributed as inverted exponentiated Rayleigh model, the author have developed estimation procedures for the stress–strength reliability parameter R = P(X>Y) under Type II hybrid censored samples.

The maximum likelihood and Bayesian estimates of R based on Type II hybrid censored samples are evaluated. Because there is no closed form for the Bayes estimate, the author use the Metropolis–Hastings algorithm to obtain approximate Bayes estimate of the reliability parameter. Furthermore, the author construct the asymptotic confidence interval, bootstrap confidence interval and highest posterior density (HPD) credible interval for R. The Monte Carlo simulation study has been conducted to compare the performance of various proposed point and interval estimators. Finally, the validity of the stress–strength reliability model is demonstrated via a practical case.

The performance of various point and interval estimators is compared via the simulation study. Among all proposed estimators, Bayes estimators using MHG algorithm show minimum MSE for all considered censoring schemes. Furthermore, the real data analysis indicates that the splashing diameter decreases with the increase of MPa under different hybrid censored samples.

The frequentist and Bayesian methods are developed to estimate the associated parameters of the reliability model under the hybrid censored inverted exponentiated Rayleigh distribution. The application of the proposed stress–strength reliability model will help the reliability engineers and also other scientists to estimate the system reliability.

The potential for overall efficiency improvements of modern hydro power turbines is a few percent. A significant part of the losses occurs in the draft tube. To improve the efficiency by analysing the flow in the draft tube, it is therefore necessary to do this accurately, i.e. one must know how large the iterative and the grid errors are. This was done by comparing three different methods to estimate errors. Four grids (122,976 to 4,592 cells) and two numerical schemes (hybrid differencing and CCCT) were used in the comparison. To assess the iterative error, the convergence history and the final value of the residuals were used. The grid error estimates were based on Richardson extrapolation and least square curve fitting. Using these methods we could, apart from estimate the error, also calculate the apparent order of the numerical schemes. The effects of using double or single precision and changing the under relaxation factors were also investigated. To check the grid error the pressure recovery factor was used. The iterative error based on the pressure recovery factor was very small for all grids (of the order 10^–4 percent for the CCCT scheme and 10^–10percent for the hybrid scheme). The grid error was about 10 percent for the finest grid and the apparent order of the numerical schemes were 1.6 for CCCT (formally second order) and 1.4 for hybrid differencing (formally first order). The conclusion is that there are several methods available that can be used in practical simulations to estimate numerical errors and that in this particular case, the errors were too large. The methods for estimating the errors also allowed us to compute the necessary grid size for a target value of the grid error. For a target value of 1 percent, the necessary grid size for this case was computed to 2 million cells.

The purpose of this paper is to introduce the analysis and performance of a hybrid multiple access scheme which combines interleave division multiple access (IDMA) and code division multiple access (CDMA), referred to as IDMA/CDMA. With experimentations, the scheme can achieve good performance with simple user ends for the system.

IDMA technique is employed as the uplink transmission and CDMA technique as the downlink transmission. The performance comparison of IDMA and IDMA/CDMA downlink is studied with Monte Carlo simulations to investigate the bit error rates. For IDMA/CDMA downlink, the signals are interleaved by random interleavers, spreaded by M‐sequence and orthogonal Gold sequence, respectively, and then transmitted over an AWGN with BPSK modulation. Moreover, a channel estimation approach for IDMA based on expectation maximization (EM) algorithm is also presented.

The multi‐user detection (MUD) algorithm in uplink IDMA systems is much simpler than that used in traditional CDMA systems. It is also well known that the orthogonal spreading codes in CDMA can effectively help suppress the MAI in the synchronized environments. But for the asynchronized application, the performance of CDMA systems will degrade due to the serious multiple user interference. According to traditional IDMA and orthogonal code division multiple access/IDMA signal detection algorithms, every UE has to detect all of the signals of other users for iterative detection. The advantages of IDMA and CDMA technologies can be utilized substantially.

The proposed hybrid multiple access scheme can achieve a very simple chip‐by‐chip iterative MUD strategy at base station, and the simplified receiving operation at UE. On the other hand, this paper also evaluates channel estimation approach for IDMA based on EM algorithm.

Large sparse least-squares problems arise in different scientific disciplines such as optimization, data analysis, machine learning and simulation. This paper aims to propose a two-level hybrid direct-iterative scheme, based on novel block independent column reordering, for efficiently solving large sparse least-squares linear systems.

Herewith, a novel block column independent set reordering scheme is used to separate the columns in two groups: columns that are block independent and columns that are coupled. The permutation scheme leads to a two-level hierarchy. Using this two-level hierarchy, the solution of the least-squares linear system results in the solution of a reduced size Schur complement-type square linear system, using the preconditioned conjugate gradient (PCG) method as well as backward substitution using the upper triangular factor, computed through sparse Q-less QR factorization of the columns that are block independent. To improve the convergence behavior of the PCG method, the upper triangular factor, computed through sparse Q-less QR factorization of the coupled columns, is used as a preconditioner. Moreover, to further reduce the fill-in, then the column approximate minimum degree (COLAMD) algorithm is used to permute the block consisting of the coupled columns.

The memory requirements for solving large sparse least-squares linear systems are significantly reduced compared to Q-less QR decomposition of the original as well as the permuted problem with COLAMD. The memory requirements are reduced further by choosing to form larger blocks of independent columns. The convergence behavior of the iterative scheme is improved due to the chosen preconditioning scheme. The proposed scheme is inherently parallel due to the introduction of block independent column reordering.

The proposed scheme is a hybrid direct-iterative approach for solving sparse least squares linear systems based on the implicit computation of a two-level approximate pseudo-inverse matrix. Numerical results indicating the applicability and effectiveness of the proposed scheme are given.

A new design framework for crossover operator is proposed based on the commonality concept. In the reproduction process the resulting hybrid crossover operator includes a local search scheme aiming to improve the genetic characteristics of the offspring. Commonality suggests that search should be driven in the neighbourhood of parents, and local optimisers can drive this search. The ranking of the offspring candidates is based on a local fitness function using approximations and appropriated heuristics linked to the structural optimisation problem. The goal of this approach is to identify and preserve the common schema of the two parents responsible for their high‐observed fitness. The proposed hybrid crossover operator is embedded into a genetic algorithm supported by an elitist strategy and its performance is compared with the parametrised uniform crossover.