认知行为计算模型结合DM的教学质量提升

为了根据认知及行为表现区分不同类别的学生以更好地提升教师教学质量,提出了基于认知行为计算模型的数据挖掘模型。基于各种重要的认知、行为输入参数,提出了认知、行为指数因子计算模型;依据所搜集的六个认知参数及三个行为参数,运用人工神经网络、灵敏度分析、数据挖掘及分类回归树算法对数据进行分类;将学生划分成三种不同的类别,从而更好地针对不同类别的学生实施不同的教学策略。实验结果表明,学生分类问题中,行为参数远比认知参数重要,分析结果表明了所提模型在教育系统教师工作支持领域的可行性。     

基于30°角同心圆环形取样的室内人员检测方法

在现有基于信道状态信息的室内无源定位方法中,取样点的选取对指纹库的特征匹配准确率以及定位精度具有较大影响.根据WiFi信号的传输特性和信道的衰落特征,提出一种30°角同心圆环形取样法.离线阶段,按照同心圆对检测区域实现环形划分并每隔30.进行一次取样,运用主成分分析算法提取差异化信号特征并构建指纹库.在线阶段,通过陆地移动距离算法进行入侵检测,当检测到有人存在时,利用改进的支持向量回归算法并引入高斯核函数对数据进行特征匹配,最终实现人员的精确定位.实验结果表明,与CSI-MIMO、FIFS方法相比,该方法定位精度更高,定位误差更小.     

Addressing the ability of a land biosphere model to predict key biophysical vegetation characterisation parameters with Global Sensitivity Analysis

Sensitivity Analysis (SA) of the SimSphere Soil Vegetation Atmosphere Transfer (SVAT) model has been performed in this study using a cutting edge and robust Global Sensitivity Analysis (GSA) approach, based on the use of the Gaussian Emulation Machine for Sensitivity Analysis (GEM-SA) tool. The sensitivity of the following model outputs was evaluated: the ambient CO₂ concentration, the rate of CO₂ uptake by the plant, the ambient O₃ concentration, the flux of O₃ from the air to the plant/soil boundary and the flux of O₃ taken up by the plant alone. The most sensitive model inputs for the majority of outputs were: The Leaf Area Index (LAI), Fractional Vegetation Cover (Fr), Cuticle Resistance (CR) and Vegetation Height (VH). The influence of the external CO₂ on the leaf and O₃ concentration in the air as input parameters was also significant. Our study provides an important step forward in the global efforts towards SimSphere verification given the increasing interest in its use as an independent modelling or educational tool. Results of this study are also timely given the ongoing global efforts focused on deriving, at an operational level, spatio-temporal estimates of energy fluxes and soil moisture content using SimSphere synergistically with Earth Observation (EO) data.     

A long-term sensitivity analysis of the denitrification and decomposition model

Although sensitivity analysis (SA) was conducted on the DeNitrification–DeComposition (DNDC) model, a global SA over a long period of time is lacking. We used a method of Bayesian analysis of computer code outputs (BACCO) with the Gaussian emulation machine for sensitivity analysis software (GEM-SA) to conduct a long-term SA of DNDC for predicting the annual change of soil organic carbon (dSOC), nitrous oxide emission (N₂O) and grain yield of spring wheat. Twenty seven non-weather input parameters with wide ranges were selected for SA using weather data recorded from Three Hills, Alberta over 86 years (1921–2006). The SA had two steps: 1) a preliminary BACCO GEM-SA was conducted to identify a more accurate emulator sampling method and to screen out parameters with insignificant influence on model outcomes; and 2) final BACCO GEM-SA was conducted with optimal input design set for emulator training runs varying only the significant input parameters. Results indicated that the Maximin Latin Hypercube sampling method outperformed the LP-τ method with higher emulator accuracy. Most of the 27 input parameters contributed little to the three outputs by the first step BACCO GEM-SA. In the second step of BACCO GEM-SA there were only three (in the case of dSOC) and six (in the cases of N₂O and yield) input parameters whose influence contributed to more than 10% of the total output variances by their total effects. Among the selected parameters, initial soil organic carbon and clay content are very important and were important in determining results for all three outputs. Sensitivities of some parameters, such as clay content and urea fertilizer amount changed dramatically over the years. This indicates that a single year SA may overestimate or underestimate a long-term parameter effect on the model prediction. The two-step procedure with the BACCO GEM-SA method improved the accuracy of SA and provided important information for model validation and parameterization.     

Global Sensitivity Analysis of environmental models: Convergence and validation

We address two critical choices in Global Sensitivity Analysis (GSA): the choice of the sample size and of the threshold for the identification of insensitive input factors. Guidance to assist users with those two choices is still insufficient. We aim at filling this gap. Firstly, we define criteria to quantify the convergence of sensitivity indices, of ranking and of screening, based on a bootstrap approach. Secondly, we investigate the screening threshold with a quantitative validation procedure for screening results. We apply the proposed methodologies to three hydrological models with varying complexity utilizing three widely-used GSA methods (RSA, Morris, Sobol’). We demonstrate that convergence of screening and ranking can be reached before sensitivity estimates stabilize. Convergence dynamics appear to be case-dependent, which suggests that “fit-for-all” rules for sample sizes should not be used. Other modellers can easily adopt our criteria and procedures for a wide range of GSA methods and cases.     

A comparative assessment of ensemble learning for credit scoring

Both statistical techniques and Artificial Intelligence (AI) techniques have been explored for credit scoring, an important finance activity. Although there are no consistent conclusions on which ones are better, recent studies suggest combining multiple classifiers, i.e., ensemble learning, may have a better performance. In this study, we conduct a comparative assessment of the performance of three popular ensemble methods, i.e., Bagging, Boosting, and Stacking, based on four base learners, i.e., Logistic Regression Analysis (LRA), Decision Tree (DT), Artificial Neural Network (ANN) and Support Vector Machine (SVM). Experimental results reveal that the three ensemble methods can substantially improve individual base learners. In particular, Bagging performs better than Boosting across all credit datasets. Stacking and Bagging DT in our experiments, get the best performance in terms of average accuracy, type I error and type II error.     

Comparison of variance-based and moment-independent global sensitivity analysis approaches by application to the SWAT model

Global Sensitivity Analysis (GSA) is an essential technique to support the calibration of environmental models by identifying the influential parameters (screening) and ranking them.In this paper, the widely-used variance-based method (Sobol') and the recently proposed moment-independent PAWN method for GSA are applied to the Soil and Water Assessment Tool (SWAT), and compared in terms of ranking and screening results of 26 SWAT parameters. In order to set a threshold for parameter screening, we propose the use of a “dummy parameter”, which has no influence on the model output. The sensitivity index of the dummy parameter is calculated from sampled data, without changing the model equations. We find that Sobol' and PAWN identify the same 12 influential parameters but rank them differently, and discuss how this result may be related to the limitations of the Sobol' method when the output distribution is asymmetric.     

Numeric sensitivity analysis applied to feedforward neural networks

During the last 10 years different interpretative methods for analysing the effect or importance of input variables on the output of a feedforward neural network have been proposed. These methods can be grouped into two sets: analysis based on the magnitude of weights; and sensitivity analysis. However, as described throughout this study, these methods present a series of limitations. We have defined and validated a new method, called Numeric Sensitivity Analysis (NSA), that overcomes these limitations, proving to be the procedure that, in general terms, best describes the effect or importance of the input variables on the output, independently of the nature (quantitative or discrete) of the variables included. The interpretative methods used in this study are implemented in the software program Sensitivity Neural Network 1.0, created by our team.     

遥感图像融合定量评价方法及实验研究

近年来融合技术不断创新,方法多种多样,但是对融合结果缺乏统一的定量评价方法。在分析与总结当前常用的遥感图像融合结果定量评价方法的基础上,给出了亮度信息、空间细节信息和光谱信息等定量评价参数,并且通过程序实现了定量评价。以IHS和PCA两种融合方法对QuickBird图像多光谱波段和全色波段融合作为实验,进行定量评价分析,结果表明所提出的定量评价参数能够较准确地反映图像融合情况,比定性评价有效、全面,可为选择恰当的融合方法提供依据。     

正交及不相关边界邻域保持嵌入的人脸识别

邻域保持嵌入(NPE)算法本质上仍是一种无监督方法,并没有有效利用已有的类别信息提高分类效率。为此提出两种有监督流形学习方法：正交边界邻域保持嵌入(OMNPE)和不相关边界邻域保持嵌入(UMNPE)。首先构造类内和类间邻接图,并定义类内和类间重构误差;然后分别在正交和不相关约束条件下寻找最小化类内重构误差同时最大化类间重构误差的投影向量;将训练样本和测试样本分别投影到低维子空间中,再利用最近邻分类器进行分类识别。在ORL和Yale人脸库上的实验结果表明,与线性判别分析(LDA)、边界Fisher分析(MFA)等子空间人脸识别算法相比,所提算法的平均识别率提高了0.5%~3%,验证了算法的有效性。     

Application of a fuzzy neural network combined with a chaos genetic algorithm and simulated annealing to short-term load forecasting

A fuzzy neural network combined with a chaos-search genetic algorithm (CGA) and simulated annealing (SA), hereafter called the FCS method, or simply the FCS, applied to short-term power-system load forecasting as a sample test is proposed in this paper. A fuzzy hyperrectangular composite neural network (FHCNN) is adopted for the initial load forecasting. An integrated CGA and fuzzy system (CGF) and SA is then used to find the optimal FHCNN parameters instead of the ones with the back propagation method. The CGF method will generate a set of parameters for a feasible solution. The CGF method holds good global search capability but poor local search ability. On the contrary, the SA method possesses a good local optimal search capability. We hence propose in this paper to combine the two methods to exploit their advantages and, furthermore, to eliminate the known downside of the traditional artificial neural network. The proposed FCS is next applied to power-system load forecasting as a sample test, which demonstrates an encouraging degree of accuracy superior to other commonly used forecasting methods available. The forecasting results are tabulated and partially converted into bar charts for evaluation and clear comparisons.     

A generic approach for the solution of nonlinear residual equations. Part III: Sensitivity computations

Sensitivity analysis provides qualitative and quantitative information on the behaviour of the model under study, and offers an access to gradients that may be used for identification purposes. Such precious information may be obtained at a low development cost applying a generic automatic differentiation (AD) tool to the computer code implementing this model. Nonlinear residual problems solved through a path following method may be addressed too. In this paper, AD techniques are adapted to the Taylor-based asymptotic numerical method. A sensitivity study of a laminated glass beam to the perturbation of some material and geometric parameters, and the perturbation of elementary stiffness matrices illustrates the method.     

Parameter sensitivity analysis of crop growth models based on the extended Fourier Amplitude Sensitivity Test method

Sensitivity analysis (SA) has become a basic tool for the understanding, application and development of models. However, in the past, little attention has been paid to the effects of the parameter sample size and parameter variation range on the parameter SA and its temporal properties. In this paper, the corn crop planted in 2008 in the Yingke Oasis of northwest China is simulated based on meteorological observation data for the inputs and statistical data for the parameters. Furthermore, using the extended Fourier Amplitude Sensitivity (EFAST) algorithm, SA is performed on the 47 crop parameters of the WOrld FOod STudies (WOFOST) crop growth models. A deep analysis is conducted, including the effects of the parameter sample size and variation range on the parameter SA, the temporal properties and the multivariable output issues of SA. The results show that sample size highly affects the convergence of the sensitivity indices. Two types of parameter variation ranges are used for the analysis, and the results show that the sensitive parameters of the two parameter spaces are distinctly different. In addition, taking the storage organ biomasses at the different growth stages as the objective output, the time-dependent characteristics of the parameter sensitivity are discussed. The results show that several sensitive parameters exist in the grain biomass throughout the entire development stage. In addition, analyzing the twelve sensitive parameters has proven that although certain parameters have no effect on the final yield, they play key roles in certain growth stages, and the importance of these parameters gradually increases. Finally, the sensitivity analyses of different state variable outputs are performed, including the biomass, yield, leaf area index, and transpiration coefficient. The results suggest that the sensitive parameters of various variable processes differ. This study highlights the importance of considering multiple characteristics of the model parameters and the responses of the models in specific phenological stages.     

基于FMEA和FTA的ZPW-2000A轨道电路可靠性分析

针对ZPW-2000A型无绝缘移频轨道电路系统的可靠性问题,提出采用故障模式影响分析（FMEA）和故障树分析（FTA）相结合的方法,对系统进行可靠性研究和分析。通过对系统分析和定义,建立故障模式影响分析表,找出所有可能的故障模式、故障后果、故障检测方法和补救措施等,在此基础上建立系统故障树,求取最小割集,进行定性和定量分析。定性分析判定系统的薄弱环节,定量分析计算顶事件的故障概率、各最小割集的重要度及系统的可靠性指标,通过与相关技术规定比较,验证了该可靠性分析方法的有效性。     

A dependability profile within MARTE

The importance of assessing software non-functional properties (NFP) beside the functional ones is well accepted in the software engineering community. In particular, dependability is a NFP that should be assessed early in the software life-cycle by evaluating the system behaviour under different fault assumptions. Dependability-specific modeling and analysis techniques include for example Failure Mode and Effect Analysis for qualitative evaluation, stochastic Petri nets for quantitative evaluation, and fault trees for both forms of evaluation. Unified Modeling Language (UML) may be specialized for different domains by using the profile mechanism. For example, the MARTE profile extends UML with concepts for modeling and quantitative analysis of real-time and embedded systems (more specifically, for schedulability and performance analysis). This paper proposes to add to MARTE a profile for dependability analysis and modeling (DAM). A case study of an intrusion-tolerant message service will offer insight on how the MARTE-DAM profile can be used to derive a stochastic Petri net model for performance and dependability assessment.     

Assessment of pan-sharpened very high-resolution WorldView-2 images

The fusion of multispectral (MS) images with high spatial resolution panchromatic (pan) images compensates the trade-off between spatial and spectral resolutions. The performance of fusion algorithms for different sensors is an active area of research. Therefore, with the availability of new very high-resolution (VHR) sensors, it becomes customary to evaluate the applicability of existing fusion techniques. In this study, we focused on the WorldView-2 (WV2) spaceborne sensor, which captures data in eight MS (2 m spatial resolution) bands and one pan (0.5 m spatial resolution) band. We compared and assessed 12 fusion techniques, namely Brovey transform (BT), Ehlers, Gram–Schimdt (GS), hyperspherical colour sphere (HCS), high-pass filter (HPF), modified intensity hue saturation (ModIHS), Multiplicative, PANSHARP, PANSHARP2, principal component (PC), and wavelet (IHS and PC)-based methods. To measure the quality of fused products, qualitative and quantitative methods were used. In qualitative methods, visual analysis of different colour composites was carried out. In quantitative methods, band-wise eight quality metrics including the required processing time and controlling parameters were reported. For overall image quality assessment, it is necessary to combine the values of different quality metrics. Therefore, a mean observation score based on these values was calculated to rank the fusion methods. It was observed that the HPF and PANSHARP methods produced the most visually appealing images, whereas quantitative values indicated that HCS, HPF, and PANSHARP methods performed better than other methods. Combining the results of quantitative and qualitative analyses, it was found that PANSHARP and HPF methods are superior to other methods in preserving both spatial and spectral details.     

Sensitivity analysis of environmental models: A systematic review with practical workflow

Sensitivity Analysis (SA) investigates how the variation in the output of a numerical model can be attributed to variations of its input factors. SA is increasingly being used in environmental modelling for a variety of purposes, including uncertainty assessment, model calibration and diagnostic evaluation, dominant control analysis and robust decision-making. In this paper we review the SA literature with the goal of providing: (i) a comprehensive view of SA approaches also in relation to other methodologies for model identification and application; (ii) a systematic classification of the most commonly used SA methods; (iii) practical guidelines for the application of SA. The paper aims at delivering an introduction to SA for non-specialist readers, as well as practical advice with best practice examples from the literature; and at stimulating the discussion within the community of SA developers and users regarding the setting of good practices and on defining priorities for future research.     

A comparative study of different machine learning methods for landslide susceptibility assessment: A case study of Uttarakhand area (India)

Landslide susceptibility assessment of Uttarakhand area of India has been done by applying five machine learning methods namely Support Vector Machines (SVM), Logistic Regression (LR), Fisher's Linear Discriminant Analysis (FLDA), Bayesian Network (BN), and Naïve Bayes (NB). Performance of these methods has been evaluated using the ROC curve and statistical index based methods. Analysis and comparison of the results show that all five landslide models performed well for landslide susceptibility assessment (AUC = 0.910–0.950). However, it has been observed that the SVM model (AUC = 0.950) has the best performance in comparison to other landslide models, followed by the LR model (AUC = 0.922), the FLDA model (AUC = 0.921), the BN model (AUC = 0.915), and the NB model (AUC = 0.910), respectively.     

Towards a unified framework for interval based qualitative computational matrix structural analysis

Qualitative reasoning about physical system behavior is one of the things engineers do to assess the “reasonableness” of the outputs from their numerical simulations, but such reasoning is typically assumed to be an exclusively human activity under the rubric of “engineering judgement”. It is the authors' contention that such reasoning can be rendered in a computational manner and that Interval Analysis (IA) provides the mathematical foundations to do so.Using Interval Analysis, the notion of operations with numbers is replaced with operations on sets of numbers. Representation of structural parameters with sets of known values permits a simultaneous parametric study. Using concepts from Qualitative and Order of Magnitude Reasoning, interval methods are augmented to perform levels of both qualitative and quantitative mechanical reasoning. Whereby interval techniques employ tight numerical intervals, it is possible to use large or even infinite intervals in order to apply qualitative computational techniques. A logic based implementation facilitated prototyping as well as approaching the unification of the various forms of analysis. Examples discussed in this paper will demonstrate that such a link between qualitative and quantitative techniques exists within a framework of interval analysis.     

High-resolution satellite image fusion using regression kriging

Image fusion is an important component of digital image processing and quantitative image analysis. Image fusion is the technique of integrating and merging information from different remote sensors to achieve refined or improved data. A number of fusion algorithms have been developed in the past two decades, and most of these methods are efficient for applications especially for same-sensor and single-date images. However, colour distortion is a common problem for multi-sensor or multi-date image fusion. In this study, a new image fusion method of regression kriging is presented. Regression kriging takes consideration of correlation between response variable (i.e., the image to be fused) and predictor variables (i.e., the image with finer spatial resolutions), spatial autocorrelation among pixels in the predictor images, and the unbiased estimation with minimized variance. Regression kriging is applied to fuse multi-temporal (e.g., Ikonos, QuickBird, and OrbView-3) images. The significant properties of image fusion using regression kriging are spectral preservation and relatively simple procedures. The qualitative assessments indicate that there is no apparent colour distortion in the fused images that coincides with the quantitative checks, which show that the fused images are highly correlated with the initial data and the per-pixel differences are too small to be considered as significant errors. Besides a basic comparison of image fusion between a wavelet based approach and regression kriging, general comparisons with other published fusion algorithms indicate that regression kriging is comparable with other sophisticated techniques for multi-sensor and multi-date image fusion.