A Grey System Theory‐Based Default Prediction Model for Construction Firms

As the prediction of construction firm failure is of great importance for owners, contractors, investors, banks, insurance firms, and creditors, previous studies have developed several models for predicting the probability of construction firm default based on financial ratio analysis. However, to be applied, these models require a considerable quantity of data, including normally distributed data, and the models cannot tolerate too many changing factors. Furthermore, most of the approaches produce sample selection biases. To avoid these disadvantages, this study is the first to integrate the grey system theory with all available firm‐year samples during the sample period to provide a new method for predicting the probability of construction firm default. This method not only offers an improved rate of prediction accuracy, but it also offers simpler and clearer procedures as a reference for examining firm default probability and ranks all financial ratios in terms of their level of importance. The research collects and analyzes the financial reports of 92 construction firms in the United States. The proposed model includes only eight ranked variables (financial ratios), and it achieves an 84.8% level of accuracy for predicting construction firm default probability. As a result, practitioners may directly use the model as a means of quickly and conveniently examining their firm default probability with the simple procedures.     

Real-time prediction of rock mass classification based on TBM operation big data and stacking technique of ensemble learning

Real-time prediction of the rock mass class in front of the tunnel face is essential for the adaptive adjustment of tunnel boring machines(TBMs).During the TBM tunnelling process,a large number of operation data are generated,reflecting the interaction between the TBM system and surrounding rock,and these data can be used to evaluate the rock mass quality.This study proposed a stacking ensemble classifier for the real-time prediction of the rock mass classification using TBM operation data.Based on the Songhua River water conveyance project,a total of 7538 TBM tunnelling cycles and the corresponding rock mass classes are obtained after data preprocessing.Then,through the tree-based feature selection method,10 key TBM operation parameters are selected,and the mean values of the 10 selected features in the stable phase after removing outliers are calculated as the inputs of classifiers.The preprocessed data are randomly divided into the training set(90%)and test set(10%)using simple random sampling.Besides stacking ensemble classifier,seven individual classifiers are established as the comparison.These classifiers include support vector machine(SVM),k-nearest neighbors(KNN),random forest(RF),gradient boosting decision tree(GBDT),decision tree(DT),logistic regression(LR)and multilayer perceptron(MLP),where the hyper-parameters of each classifier are optimised using the grid search method.The prediction results show that the stacking ensemble classifier has a better performance than individual classifiers,and it shows a more powerful learning and generalisation ability for small and imbalanced samples.Additionally,a relative balance training set is obtained by the synthetic minority oversampling technique(SMOTE),and the influence of sample imbalance on the prediction performance is discussed.     

基于粒子群算法优化支持向量机的岩爆预测研究

由于地下工程的复杂性,岩爆的发生受到多种因素的影响,目前尚没有一种可靠的预测方法来对其进行预报,进而有针对性地进行工程灾害的风险控制。笔者提出将应力强度比(σ_θ/σ_c)、脆性系数(σ_c/σ_t)和弹性能量指数(Wet)作为影响岩爆的主要指标,并根据粒子群优化算法的参数选取和收敛速度快的优势及支持向量机的小样本、高维度、非线性的特性,提出了用粒子群优化算法对影响支持向量机分类性能的两个主要参数进行优化,进而获得优化的支持向量机分类器。利用PSO-SVM对在建二广九标茅田界隧道深埋变质砂岩岩爆发生情况进行预测,定量地判断该标段不存在岩爆现象,预测结果与茅田界隧道的实际情况基本相符。     

概率积分法参数辨识的多尺度核偏最小二乘回归方法

 针对传统的偏最小二乘回归(PLS)、人工神经网络(ANN)、支持向量机(SVM)等非线性建模方法在概率积分法参数辨识中存在着预测效果差的不足,提出概率积分法参数辨识的多尺度核偏最小二乘回归(multi-scale KPLS)方法。首先,构建满足容许条件的多尺度高斯核函数;然后,对学习样本进行模糊聚类,以最优分类个数作为多尺度高斯核函数的尺度个数,并采用10次10折交叉验证按照网格搜索方法确定核函数的宽度;最后,详细论述multi-scale KPLS的建模过程。通过实例将multi-scale KPLS的预测结果与3种传统的PLS方法、径向基神经网络(RBF-NN)和SVM模型进行对比分析。结果表明：multi-scale KPLS顾及建模样本的多尺度特性,其预测精度明显高于其他预测模型;multi-scale KPLS有效地克服了各影响因素之间的多重共线性对预测结果的不利影响,具有较强的稳健性;multi-scale KPLS适用于多个因变量对多个自变量的概率积分法参数辨识问题,其建模参数均可自适应确定,在建模效率上优于RBF-NN和SVM。     

Ground-level ozone prediction by support vector machine approach with a cost-sensitive classification scheme

For ground-level ozone (O(3)) prediction, a predictive model, with reliable performance not only on non-polluted days but, more importantly, on polluted days, is favored by public authorities to issue alerts, so that concerned citizens and industrial organizations could take precautions to avoid exposure and reduce harmful emissions. However, the class imbalance problem, i.e., in some collected field data, number of O(3) polluted days are much smaller than that of non-polluted days, will deteriorate the model performance on minority class-O(3) polluted days. Despite support vector machine (SVM) obtaining promising results in air quality prediction, in this study, a cost-sensitive classification scheme is proposed for the standard support vector classification model (S-SVC) in order to investigate whether the class imbalance plagues S-SVC. The S-SVC with such scheme is named as CS-SVC. Experiments on imbalanced data sets collected from two air quality monitoring sites in Hong Kong show that 1) S-SVC is still sensitive to class imbalance problem; 2) compared with S-SVC, CS-SVC effectively avoids class imbalance problem with lower percentage of false negative on O(3) polluted days but with higher percentage of false positive on non-polluted days; 3) compared with both S-SVC and CS-SVC, support vector regression model (SVR), after converting its output to binary one, only has similar performance with S-SVC, which indicates class imbalance problem also impairs the regressor model. From point of protecting public health, CS-SVC, which less likely misses to forecast O(3) polluted days, is recommended here.     

Predicting profitability of listed construction companies based on principal component analysis and support vector machine—Evidence from China

In order to monitor the operating conditions of the construction industry, this paper incorporates the principal component analysis (PCA) and support vector machine (SVM) to predict the profitability of the construction companies listed on A-share market in China. With annual financial data in 2001–2012, this paper selected six indicators from different profitable perspectives to build a composite profitability index based on the PCA technique, and then established a SVM model to make the corporate profitability prediction of the construction companies in China. The results indicate that, the technical combination of the PCA and SVM can improve the profitability prediction significantly. In 2003–2012, the accuracy of predicting the profitability of the Chinese construction companies exceeded 80% on average. Compared with the artificial neural network (ANN), the SVM model has the superiority in the accuracy prediction of the Chinese construction companies.     

滑坡位移预测的支持向量机模型参数选择研究

支持向量机(Support Vector Machine, SVM)已被广泛应用到滑坡位移预测,但在具体使用时,SVM的惩罚系数C、核函数参数δ及松弛系数ζ这三个重要参数的取值选择成为影响预测精度的关键。为有效分析SVM三参数取值对滑坡位移预测精度的影响规律,以三峡库区浮托减重和动水压力型两类典型水库滑坡为代表的连续6年地表位移、降雨及库水位监测数据为研究对象,首先,采用移动平均法将位移数据分解为趋势项和周期波动项,并区分训练集和检验集;再结合对滑坡变形机理及影响因素的分析,选择相应预测变量分别建立趋势项和波动项位移预测SVM模型;然后,在固定两参数情形下,通过改变另一参数的取值大小以获得SVM训练集与检验集的预测精度变化规律;最后,建立起典型水库滑坡SVM位移分解预测的参数取值推荐范围。该取值范围可以作为滑坡位移预测SVM模型的参数寻优初始搜索范围,可以在保证预测精度的前提下大大提高搜索效率。     

Slope stability prediction using ensemble learning techniques: A case study in Yunyang County,Chongqing, China

Slope stability prediction plays a significant role in landslide disaster prevention and mitigation. This study develops an ensemble learning-based method to predict the slope stability by introducing the random forest(RF) and extreme gradient boosting(XGBoost). As an illustration, the proposed approach is applied to the stability prediction of 786 landslide cases in Yunyang County, Chongqing, China. For comparison, the predictive performance of RF, XGBoost, support vector machine(SVM), and logi...     

烃类物质自燃点的QSPR预测研究

应用定量构效关系(QSPR)方法对烃类物质的自燃点开展了预测研究.选取国际电工委员会数据库中的39种烃类物质作为样本集,随机选择34种作为训练集,5种作为测试集.采用遗传算法(GA)对变量进行筛选,结合线性和非线性方法分别建立多元线性回归(MLR)模型和支持向量机(SVM)模型,理论预测得到了5种烃类物质的自燃点.结果...     

基于边坡位移监测数据的进化支持向量机预测模型研究

支持向量机方法是基于统计学习理论和结构风险最小化原则的学习方法,在回归预测方面具有良好外推能力,并且适合小样本的统计学习问题。建立支持向量机预测模型,对边坡位移进行预测计算,将预测值和实测值对比分析,验证了支持向量机预测模型较强的外推能力和预测计算的有效性。通过对边坡位移初始时序位移数据进行灰色理论的累加生成和累减生成处理,形成新的时间序列数据,在此基础上,计算出预测值,并与基于初始时间序列的支持向量机预测结果对比分析,基于新生成的时间序列数据进行预测计算结果精度明显提高。基于边坡位移监测数据构建训练样本数据集,研究了训练样本数据集的选取对预测结果的影响。对支持向量机预测模型的关键参数进行敏感度分析,并采用进化算法–微粒群算法对支持向量机模型参数加以优化,提高了预测精度。     

Developing an SVM based risk hedging prediction model for construction material suppliers

Construction material suppliers are usually exposed to financial risks as a consequence of a high debt capital structure and the nature of the material import business. There is demand for a tool that is able to predict whether such a material supplier, based on its financial status, should use derivatives to hedge financial risks. The research objective is to develop a prediction model using the Support Vector Machine (SVM) to determine whether employing risk hedging based on derivatives usage would be beneficial. The scope of this research limits the database to 640 financial statements published over the last 5 years from 32 listed construction material suppliers. A total of 10 input determinants were identified and verified from the literature review, t-test results, and collinearity diagnosis. Using data trimming and normalization, these 640 sets were downsized to 520 sets which contained 248 effective and 272 ineffective risk-hedging sets. The SVM prediction model, based on the kernel radial basis function and normalized data, yields a prediction accuracy rate of 80.65%. The evaluation, using logistics and small sets of data, shows the validation and practicality of this model. This research concludes that 10 financial determinates are proven candidates for financial risk hedging. From the viewpoint of derivatives usage and the proposed SVM prediction model it appears feasible for construction material suppliers to apply this model.     

烃的含氧衍生物自燃点的QSPR预测研究

应用定量构效关系（QSPR）方法对烃的含氧衍生物的自燃点（AIT）及其与分子结构间的内在定量关系进行了研究。选取国际电工委员会（IEC）数据库中的76种烃的含氧衍生物作为样本集,选择65 种作为训练集用于建立预测模型,11 种作为测试集。采用遗传算法（GA）对变量进行筛选,结合线性和非线性方法分别建立多元线性回归( MLR) 模型和支持向量机( SVM) 模型,理论预测得到了11种烃的含氧衍生物的自燃点,最后对所构建模型的性能及应用域进行了评价。结果表明,经GA筛选得出MATS2e、nCOH、Dv、BEHv2、nCHR、GATS1v、IDE、Du等8种特征分子描述符,GA-MLR和GA-SVM模型的理论预测值与实验值均较为相符且后者更优,两个预测模型均比较稳定,且具备较强的预测能力和泛化推广性能。     

基于自注意力的隧道视频火灾识别技术研究

针对公路隧道火灾样本量少、深度学习效果不理想的问题，研究一种小样本学习技术，以提高对隧道火灾样本的利用率，并在此基础上利用成熟的机器学习方法，提出一种基于自注意力的隧道视频火灾识别技术。该技术采用自注意力机制结合SVM分类器搭建火焰识别模型，该模型针对各项特征对火焰识别的重要性分配不同的注意力权重，形成注意力矩阵，并将权重矩阵与特征向量的值相加权，通过SVM的Hinge Loss进行线性支持向量机分类，对公路隧道火灾进行识别和预警。在火灾识别训练过程中，通过对火焰疑似区域进行检测，并利用数据增强技术达到样本扩增的目的，随后采用多通道融合的特征提取方式构建特征向量，输入设计的自注意力火焰识别模型中，通过梯度下降优化器进行小批量模型训练，降低迭代次数，最终获得最优特征权重参数，得到最佳识别模型。试验结果表明，该方法在模型训练时收敛较快，在火焰识别时相比未使用小样本学习的传统SVM算法，准确率提高了5%，因此能在小样本环境下有效提高火灾识别的准确度。     

Modeling of bentonite/sepiolite plastic concrete compressive strength using artificial neural network and support vector machine

Plastic concrete is an engineering material, which is commonly used for construction of cut-off walls to prevent water seepage under the dam. This paper aims to explore two machine learning algorithms including artificial neural network (ANN) and support vector machine (SVM) to predict the compressive strength of bentonite/sepiolite plastic concretes. For this purpose, two unique sets of 72 data for compressive strength of bentonite and sepiolite plastic concrete samples (totally 144 data) were prepared by conducting an experimental study. The results confirm the ability of ANN and SVM models in prediction processes. Also, Sensitivity analysis of the best obtained model indicated that cement and silty clay have the maximum and minimum influences on the compressive strength, respectively. In addition, investigation of the effect of measurement error of input variables showed that change in the sand content (amount) and curing time will have the maximum and minimum effects on the output mean absolute percent error (MAPE) of model, respectively. Finally, the influence of different variables on the plastic concrete compressive strength values was evaluated by conducting parametric studies.     

Stochastic modelling and prediction of contractor default risk

The purpose of this paper is to describe a systematic framework of stochastic modelling and prediction of financial default risk of construction contractors. Net-worth-to-asset ratio is identified as an index for default process modelling. The default condition is defined as when the ratio becomes negative the first time. A mean-reverting dynamic model for the contractor default process is found by statistical analysis and is justified by using the theory of optimal capital structure. The stochastic modelling of default uses the time to default as the fundamental random variable. A discrete time trinomial Markov chain model is developed to assess default risk in terms of a cumulative default probability function, a default probability function, and the mean and variance of time to default. Practical examples are given to illustrate the stochastic methods. A default discriminant study on a group of contractors and publicly traded companies validates the methods, and indicates a high predictability of events of default and declines of credit rating.     

Predicting the presence of E. coli in tap water using machine learning in Nepal

Within developing countries, a multitude of problems that affect the water supply process can result in the contamination of water taps. While machine learning applications have become popular for attaining efficient water quality predictions, acquiring the necessary data for modelling for developing countries is challenging. This study constructs water quality prediction models by machine learning with a pseudo-pipeline network to complement the missing data of the water supply process. Using both water source and water tap quality information measured by the Government of Nepal, we apply the three machine learning models: support vector machine (SVM), random forest (RF) and LightGBM. Furthermore, we also apply a traditional statistical method—logistic regression (LR)—to the prediction of the Escherichia coli (E. coli) contamination in water taps. With some input variables (such as the length from the nearest sources) obtained from the pseudo-pipeline network, the results show that SVM has stable and high accuracy for both the 26 cities (70%) and for the 25 cities except for Kathmandu (79%). LR performed a significantly lower accuracy for all cities (61%) than for 25 cities (79%). Additionally, we show that our method can be applied to other regions where a water quality survey has not yet been conducted.     

导水裂缝带高度预测的模糊支持向量机模型

针对传统支持向量机(SVM)模型在导水裂缝带高度预测中存在着易受奇异值干扰而造成的泛化能力降低问题,提出了基于异常样本探测、剔除的模糊支持向量机模型(FSVM)。采用模糊聚类分析和加权支持向量机(WSVM)相结合的方法,首先根据模糊ISODATA算法求得导水裂缝带高度及其影响因素的最优分类矩阵,剔除分类结果不一致的观测样本,然后以模糊隶属度为样本权重,按照WSVM建模思想建立了导水裂缝带高度预测的FSVM模型。通过实例将FSVM和WSVM、SVM的预测结果作对比分析。结果表明,FSVM避免了异常样本对预测结果的影响,并顾及了建模样本的不同重要程度,其预测精度比WSVM和SVM有较大的提高。     

Factorial design–machine learning approach for predicting incident durations

This research proposes a hybrid approach for predicting incident duration that integrates the salient features of both factorial design of experiments (DOE) and machine learning (ML). This study compares DOE with another widely used technique, forward sequential feature selection (FSFS). Moreover, to confirm the effectiveness and robustness of the proposed approach, multiple ML techniques are employed, including linear regression, decision trees, support vector machines, ensemble trees, Gaussian process regression, and artificial neural networks. The study results are validated using data from the Houston TranStar incidents archive with over 90,000 records. The accuracy of the developed predictive models is compared based on multiple techniques (i.e., no feature selection–ML, FSFS–ML, and DOE–ML). The results revealed that the significant factors affecting incident duration identified by both DOE and FSFS include the type of vehicles involved, type of lanes affected, number of vehicles involved, number of emergency responses dispatched, incident severity level, and day of the week. The comparative results of the different feature selection and modeling approaches revealed that the hybrid DOE–ML approach outperformed the other tested analysis approaches. The best-performing model under the DOE–ML approach was the SVM with cubic kernel model. It reduced the modeling time by 83.8% while increasing the prediction error by merely 0.02%, which is not significant. Therefore, the prediction accuracy could be slightly downgraded in return for a substantial reduction in the number of variables utilized, resulting in substantial savings in the modeling time and required dataset.     

岩爆基于支持向量机预测的智能模型

根据岩爆预测的特性,引入了支持向量机SVM的新方法,研究了支持向量机的基本原理及其在岩爆预测中的模型建立,通过某工程的实际应用证明：支持向量机在岩爆预测中取得了较好的效果。     

Probability and interval hybrid reliability analysis based on adaptive local approximation of projection outlines using support vector machine

This paper investigates structural reliability analysis with both random and interval variables, which is defined as a three‐classification problem and handled by support vector machine (SVM). First, it is determined that projection outlines on the limit‐state surface are crucial for describing separating hyperplanes of the three‐classification problem. Compared with the whole limit‐state surface, the region of projection outlines are much smaller. It will be beneficial to reduce the number of update points and the computational cost if SVM update concentrates on refining the approximate projection outlines. An adaptive local approximation method is developed to realize that the initial built SVM model is sequentially updated by adding new training samples located around the projection outlines. Using this method, the separating hyperplanes can be accurately and efficiently approximated by SVM. Finally, a new method is proposed to evaluate the failure probability interval based on Monte Carlo simulation and the refined SVM.