基于RS GA ELM的矿井瓦斯涌出量预测

为了对煤矿井下瓦斯涌出量进行预测,采用粗糙集与改进极限学习机相结合的方法,在样本数据的筛选上吸取粗糙集数据约简的优点,充分利用极限学习机训练速度快、具有良好泛化性能的特点,并结合遗传算法选择最优的输入权值矩阵和隐含层偏差,避免随机产生所造成的误差。利用编写程序确定隐含层神经元个数,比依靠经验更为准确。在实际应用中选取煤层瓦斯含量、煤层埋藏深度、煤层厚度、煤层间距、工作面日产量五个因素作为预测的影响参数。研究结果表明:该预测模型预测的最大相对误差为5687 1%,最小相对误差为0,平均相对误差为2582 7%,相比改进前的预测模型具有更强的泛化能力和更高的预测精度。     

利用MODIS反演武汉东湖叶绿素a浓度

以武汉东湖作为研究区域,对经过大气校正后MODIS影像的波段反射率与叶绿素a浓度实测值进行相关分析,分别应用BP人工神经网络模型和线性回归模型对武汉东湖的叶绿素a浓度进行了反演,并对两种反演方法的拟合和预测效果进行了比较.利用BP神经网络反演得到的拟合值与叶绿素a实测值的拟合效果略好于线性回归方法得到的结果,神经网络模型的可决系数R2值0.90大于线性回归模型的R2值0.820.神经网络模型预测的最小绝对误差为0.07 μg/L,线性回归模型的最小绝对误差为2.08 μg/L.最后分析了两个模型各自的优势,将模型应用到武汉东湖2008年5月19日的MODIS影像上反演出东湖水体叶绿素a浓度的分布情况,并对东湖水质进行了评价,结论与多年的地面监测结果一致.     

基于Bagging神经网络集成的燃料电池性能预测方法

利用三维质子交换膜燃料电池数学模型模拟研究了电池流道进、出口高度对电池性能的影响,然后将数值模拟结果作为神经网络模型的训练数据.以流道进、出口高度和电池电压值作为输入变量,以电池电流密度作为输出变量,建立了3层反向传播神经网络模型;然后利用Bagging集成学习方法对神经网络模型进行集成,构建了燃料电池性能预测方法.研究发现:与单一神经网络模型相比, Bagging神经网络集成模型预测精度更高,且所需模型训练数据量更少.此外对于超出训练数据以外的情形, Bagging神经网络集成模型仍然能够准确地预测燃料电池的性能,且精度良好,表明Bagging神经网络集成模型的鲁棒性较好,可用于更宽工况范围内燃料电池性能的快速预测.     

基于支持向量机回归与分源预测法的瓦斯涌出量预测模型及其应用

针对目前瓦斯涌出量预测模型存在的局限性及精度低等问题,应用分源预测和支持向量机(SVM)的基本原理,将SVM回归与分源预测法相结合,并利用SVM对回采工作面的瓦斯涌出量进行回归分析和数值模拟,建立了SVM分源预测的数学模型,提出了SVM分源预测的新方法。数值实验表明,将训练成功的SVM模型对现场数据进行回归预测并对比预测结果与实际值发现,SVM比BP神经网络预测精度更高,训练样本期望输出与实际值的最大相对误差为1.45%,小于实际要求的5%,准确率较高,预测风险低,可以满足实际要求。     

顾及大坝位移残差序列混沌效应的GA-BP预测模型

针对大坝观测数据常规模型训练后的残差混沌效应及模型回归方法的拟合度等问题,文中融合遗传算法与神经网络的数据训练优势,通过构建的遗传神经网络(GA-BP)算法对大坝变形观测序列资料进行回归提取残差序列.基于位移回归残差序列的混沌特性,利用混沌理论对其残差序列进行数值分析,并将残差预测结果与GA-BP预测模型进行叠加.据此,提出了考虑大坝变形残差序列混沌效应的GA-BP监控预测模型.实例表明,文中建立的预测模型的计算精度及收敛速度均得到提高,且考虑残差影响的大坝监控模型的预测效果得到了有效的提升.该模型的建模方法亦可推广应用于边坡及其他水工建筑物的安全预警.     

基于深度强化学习的机械装备剩余使用寿命预测方法

随着机械装备发展的日益大型化和复杂化,其使用安全性和可靠性也越来越受到重视.剩余使用寿命(remaining useful life, RUL)预测技术,通过分析设备的运行数据,预测设备还能正常运行的时间,利用该技术可有效提升设备运行的安全性和可靠性,同时可为设备的后续维修决策提供关键依据.本文提出一种基于深度强化学习(deep reinforcement learning, DRL)的RUL预测方法,首先通过自编码器(autoencoder, AE)对机械装备的原始信号进行特征提取,利用提取到的特征构成强化学习的状态变量,通过设置合适的动作空间和奖励函数训练强化学习模型,使其能依据样本间的时序相关性对装备的RUL进行准确预测.与其他方法相比,强化学习方法的时序交互决策逻辑可以自然地保留样本间的时序依赖关系,降低了RUL预测的波动性.最后利用涡轮发动机数据集CMPASS对提出的方法进行实验验证,所提出的方法在RMSE和Score两项指标上均优于目前多种RUL预测方法,且所提出方法对于接近退化末期的装备预测效果有明显提升.     

质量剖面下建筑施工项目可靠性测度模型及其应用研究

根据建筑工程质量形成的一般规律,建立了包括10个节点要素的工程项目质量可靠性分析的贝叶斯网络(BN)结构;从铁路建设项目施工现场收集到数据后进行CPT参数学习,获得了完整的BN网络;采用两个典型样本数据对模型进行了检验,针对模型中的主要节点变量进行BN推理和敏感性分析,结果显示模型具有较好的预测功能。     

一类非线性回归模型的曲率降低问题

对于相当一般的一类非线性回归模型，从理论上证明了存在一种通过增加观测点降低非线性模型曲率的方法，使非线性模型有可能渐近于线性模型，这一结论对于实际应用具有重要的指导意义，并已在航天测量数据处理中取得了满意的效果。     

基于POD-Observer技术的非定常气动力建模方法

基于本征正交分解（POD）结合观测器（Observer）技术,发展了一种新的适合于气动弹性分析的非定常气动力降阶方法.通过全阶系统行为的样本采用POD方法导出一组流体模态.将POD训练的全阶响应投影到流体模态上,得到模态幅值的响应时间历程.经由deadbeat观测器处理,这些训练数据用于识别模态幅值动态系统的Markov参数.采用特征实现算法基于上述的Markov参数构建系统的状态空间模型.算例选取了亚声速流场中的二维翼型系统.结果表明降阶模型复现了全阶系统的主要动态特性,极大缩减了原系统的自由度数量并且显著提高了计算效率.     

基于一元线性回归由风速传感器测量值推导巷道平均风速

通过试验数据建立了风速传感器的测量值与平均风速之间的一元线性回归方程,并对拟合效果进行了评价。对参数的判断表明该方程是有效的,可以近似地把风速传感器的测量值转换成平均风速值,并在相同的巷道长度、相同的风流速度、不同断面下用Comsol模拟了风速流场分布情况。     

Seasonal prediction of European cereal prices: good forecasts using bad models?

Because of their natural adherence to the climate and pronounced seasonal cycles, prices of field crops constitute an interesting field for exploring seasonal time series models. We consider quarterly prices of two major cereals: barley and wheat. Using traditional in‐sample fit and moving‐window techniques, we investigate whether seasonality is deterministic or unit‐root stochastic and whether seasonal cycles have converged over time. We find that seasonal cycles in the data are mainly deterministic and that evidence on common cycles across countries differs for the two commodities. Out‐of‐sample prediction experiments, however, yield a ranking with respect to accuracy that does not match the statistical in‐sample evidence. Parametric bootstrap experiments establish that the observed mismatch is indeed an inherent and systematic feature. Copyright © 2008 John Wiley & Sons, Ltd.     

Predicting US bank failures: A comparison of logit and data mining models

Predicting bank failures is important as it enables bank regulators to take timely actions to prevent bank failures or reduce the cost of rescuing banks. This paper compares the logit model and data mining models in the prediction of bank failures in the USA between 2002 and 2010 using levels and rates of change of 16 financial ratios based on a cross‐section sample. The models are estimated for the in‐sample period 2002–2009, while data for the year 2010 are used for out‐of‐sample tests. The results suggest that the logit model predicts bank failures in‐sample less precisely than data mining models, but produces fewer missed failures and false alarms out‐of‐sample.     

Estimating the Out‐of‐Sample Predictive Ability of Trading Rules: A Robust Bootstrap Approach

In this paper, we provide a novel way to estimate the out‐of‐sample predictive ability of a trading rule. Usually, this ability is estimated using a sample‐splitting scheme, true out‐of‐sample data being rarely available. We argue that this method makes poor use of the available data and creates data‐mining possibilities. Instead, we introduce an alternative.632 bootstrap approach. This method enables building in‐sample and out‐of‐sample bootstrap datasets that do not overlap but exhibit the same time dependencies. We show in a simulation study that this technique drastically reduces the mean squared error of the estimated predictive ability. We illustrate our methodology on IBM, MSFT and DJIA stock prices, where we compare 11 trading rules specifications. For the considered datasets, two different filter rule specifications have the highest out‐of‐sample mean excess returns. However, all tested rules cannot beat a simple buy‐and‐hold strategy when trading at a daily frequency. Copyright © 2015 John Wiley & Sons, Ltd.     

A note on in‐sample and out‐of‐sample tests for Granger causality

This paper studies in‐sample and out‐of‐sample tests for Granger causality using Monte Carlo simulation. The results show that the out‐of‐sample tests may be more powerful than the in‐sample tests when discrete structural breaks appear in time series data. Further, an empirical example investigating Taiwan's investment–saving relationship shows that Taiwan's domestic savings may be helpful in predicting domestic investments. It further illustrates that a possible Granger causal relationship is detected by out‐of‐sample tests while the in‐sample test fails to reject the null of non‐causality. Copyright © 2005 John Wiley & Sons, Ltd.     

The role of age‐structured education data for economic growth forecasts

This paper utilizes for the first time age‐structured human capital data for economic growth forecasting. We concentrate on pooled cross‐country data of 65 countries over six 5‐year periods (1970–2000) and consider specifications chosen by model selection criteria, Bayesian model averaging methodologies based on in‐sample and out‐of‐sample goodness of fit and on adaptive regression by mixing. The results indicate that forecast averaging and exploiting the demographic dimension of education data improve economic growth forecasts systematically. In particular, the results are very promising for improving economic growth predictions in developing countries. Copyright © 2009 John Wiley & Sons, Ltd.     

Scenario planning: An investigation of the construct and its measurement

Scenario‐planning academicians and practitioners have been observing for more than three decades the importance of this method in dealing with environmental uncertainty. However, there has been no valid scale that may help organizational leaders to act in practice. Our review of prior studies identifies some problems related to conceptualization, reliability, and validity of this construct. We address these concerns by developing and validating a measure of scenario planning based on Churchill's paradigm (Journal of Marketing Research, 1979, 16, 64–73). Our data analysis follows from a sample of 133 managers operating in the healthcare field in France. To validate our scale, we used three approaches: first, an exploratory factor analysis; second, an examination of psychometric proprieties of all dimensions; and third, a confirmatory factor analysis. The results of this study indicate that scenario planning is a multidimensional construct composed of three dimensions: information acquisition, knowledge dissemination, and scenario development and strategic choices.     

Nowcasting and predicting data revisions using panel survey data

The qualitative responses that firms give to business survey questions regarding changes in their own output provide a real‐time signal of official output changes. The most commonly used method to produce an aggregate quantitative indicator from business survey responses—the net balance or diffusion index—has changed little in 40 years. This paper investigates whether an improved real‐time signal of official output data changes can be derived from a recently advanced method on the aggregation of survey data from panel responses. We find, in a New Zealand application, that exploiting the panel dimension to qualitative survey data gives a better in‐sample signal about official data than traditional methods. Out‐of‐sample, it is less clear that it matters how survey data are quantified, with simpler and more parsimonious methods hard to improve. It is clear, nevertheless, that survey data, exploited in some form, help to explain revisions to official data. Copyright © 2009 John Wiley & Sons, Ltd.     

Can urban coffee consumption help predict US inflation?

Motivated by the importance of coffee to Americans and the significance of the coffee subsector to the US economy, we pursue three notable innovations. First, we augment the traditional Phillips curve model with the coffee price as a predictor, and show that the resulting model outperforms the traditional variant in both in‐sample and out‐of‐sample predictability of US inflation. Second, we demonstrate the need to account for the inherent statistical features of predictors such as persistence, endogeneity, and conditional heteroskedasticity effects when dealing with US inflation. Consequently, we offer robust illustrations to show that the choice of estimator matters for improved US inflation forecasts. Third, the proposed augmented Phillips curve also outperforms time series models such as autoregressive integrated moving average and the fractionally integrated version for both in‐sample and out‐of‐sample forecasts. Our results show that augmenting the traditional Phillips curve with the urban coffee price will produce better forecast results for US inflation only when the statistical effects are captured in the estimation process. Our results are robust to alternative measures of inflation, different data frequencies, higher order moments, multiple data samples and multiple forecast horizons.     

Nelson–Siegel,Affine and Quadratic Yield Curve Specifications: Which One is Better at Forecasting?

In this paper we compare the in‐sample fit and out‐of‐sample forecasting performance of no‐arbitrage quadratic, essentially affine and dynamic Nelson–Siegel term structure models. In total, 11 model variants are evaluated, comprising five quadratic, four affine and two Nelson–Siegel models. Recursive re‐estimation and out‐of‐sample 1‐, 6‐ and 12‐month‐ahead forecasts are generated and evaluated using monthly US data for yields observed at maturities of 1, 6, 12, 24, 60 and 120 months. Our results indicate that quadratic models provide the best in‐sample fit, while the best out‐of‐sample performance is generated by three‐factor affine models and the dynamic Nelson–Siegel model variants. Statistical tests fail to identify one single best forecasting model class. Copyright © 2011 John Wiley & Sons, Ltd.     

Measuring large‐scale market responses and forecasting aggregated sales: Regression for sparse high‐dimensional data

In this article, we propose a regression model for sparse high‐dimensional data from aggregated store‐level sales data. The modeling procedure includes two sub‐models of topic model and hierarchical factor regressions. These are applied in sequence to accommodate high dimensionality and sparseness and facilitate managerial interpretation. First, the topic model is applied to aggregated data to decompose the daily aggregated sales volume of a product into sub‐sales for several topics by allocating each unit sale (“word” in text analysis) in a day (“document”) into a topic based on joint‐purchase information. This stage reduces the dimensionality of data inside topics because the topic distribution is nonuniform and product sales are mostly allocated into smaller numbers of topics. Next, the market response regression model for the topic is estimated from information about items in the same topic. The hierarchical factor regression model we introduce, based on canonical correlation analysis for original high‐dimensional sample spaces, further reduces the dimensionality within topics. Feature selection is then performed on the basis of the credible interval of the parameters' posterior density. Empirical results show that (i) our model allows managerial implications from topic‐wise market responses according to the particular context, and (ii) it performs better than do conventional category regressions in both in‐sample and out‐of‐sample forecasts.