关键词： 主成分分析法   BP神经网络   GBDT   五峰组—龙马溪组   川南  

摘要： 为了建立合理准确的川南五峰组—龙马溪组页岩TOC含量预测方法,以长宁、泸州等地区的测井曲线及17口井实测TOC含量数据为基础,利用主成分分析法对这些资料进行预处理,基于BP神经网络和梯度提升决策树(GBDT)方法建立2种TOC含量预测模型,并将之与传统TOC含量预测方法进行对比。结果表明:(1) 2种新模型的准确度均高于传统方法,预测结果与实际值吻合度均满足要求。(2)与BP神经网络模型相比,GBDT预测精度更高,均方根误差仅为0.0387。利用GBDT方法所建立的TOC含量预测模型具有低成本、高效、连续等特点,能够快速准确地预测目的层TOC含量。该成果可为提高页岩油气勘探开发效率提供有效技术支撑。

关键词： stamp forming   machine learning   FEA   IBMLM   SBMLM   MLP   Res-SE-U-Net   CAD   CAM   CAE   design for manufacturing   modeling and simulation   sheet and tube metal forming  

摘要： In design for forming, it is becoming increasingly significant to develop surrogate models of high-fidelity finite element analysis (FEA) simulations of forming processes to achieve effective component feasibility assessment as well as process and component optimizations. However, surrogate models using traditional scalar-based machine learning methods (SBMLMs) fall short on accuracy and generalizability. This is because SBMLMs fail to harness the location information available from the simulations. To overcome this shortcoming, the theoretical feasibility and practical advantages of innovatively applying image-based machine learning methods (IBMLMs) in developing surrogate models of sheet stamp forming simulations are explored in this study. To demonstrate the advantages of IBMLMs, the effect of the location information on both design variables and simulated physical fields is first proposed and analyzed. Based on a sheet steel stamping case study, a Res-SE-U-Net IBMLM surrogate model of stamping simulations is then developed and compared with a baseline multilayer perceptron (MLP) SBMLM surrogate model. The results show that the IBMLM model is advantageous over the MLP SBMLM model in accuracy, generalizability, robustness, and informativeness. This article presents a promising methodology in leveraging IBMLMs as surrogate models to make maximum use of information from stamp forming FEA results. Future prospective studies that are inspired by this article are also discussed.

关键词： 机器学习   随机森林算法   极限学习机   支持向量机   风电功率预报  

摘要： 根据甘肃省华家岭风力发电场的风场变化特征,利用风电场2017年8月-2018年7月的风电功率监测数据及同期欧洲中期天气预报中心的数值模式预报资料,用随机森林(RF)算法分析和筛选出主要的预报因子,分别选择RF、极限学习机和支持向量机3种机器学习方法建立预报模型,通过对比预报效果,得出适合的预报方法和模型.结果表明,RF算法的平均预报均方根误差为15.6%,预报效果优于极限学习机和支持向量机(预报均方根误差分别为16.8%和17.2%).RF算法在风电功率的短期预报方面取得了更好的效果,预报值与实际监测值更加接近;基于3种机器学习算法建立的风电功率预报模型的预报结果误差值都会随着风速的增大而增大,随着风速的减小而减小.

关键词： 夏季旱涝   决策树   随机森林  

摘要： 利用1961—2010年重庆34个气象观测站夏季降水资料及国家气候中心130项环流指数,采用机器学习的决策树和随机森林方法建立重庆夏季旱涝预测模型,通过2011—2018年预测效果检验发现,夏季同期环流指数决策树模型和前冬海温指数决策树模型预测的8 a降水异常趋势均正确,比考虑单一指数的PC评分分别提高37.5%和12.5%。此外,用随机森林模型预测重庆2014—2018年的夏季降水,5 a平均PS、CC和PC评分分别是84.6、0.27和67.1,相比于业务发布预报质量均有明显提高,且随机森林的预测质量较为稳定。

关键词： 机器学习   神经网络   支持向量机   朴素贝叶斯   集成学习  

摘要： 机器学习方法在处理复杂数据、构建高精度模型方面具有显著优势,在不同领域的研究中都得到了广泛应用。数字经济时代为经济领域带来海量数据的同时,也对经济研究提出诸多挑战。机器学习方法能充分挖掘数据中的非线性、非平稳信息,有效提高经济分析结果的精度。机器学习与经济问题的融合改变了传统经济学的研究范式。文章对近年来机器学习方法在经济研究中的应用进行回顾,从通货膨胀、汇率与货币、GDP、劳动力市场、社会稳定、政策评价等角度进行总结,比较了常用机器学习方法的优缺点,并展示了模型的评价准则,如均方根误差、F1-得分、AUC值等。

关键词： L_(2)惩罚逻辑回归   支持向量机   梯度提升树算法   人工神经网络   极限梯度提升算法  

摘要： 肝癌在所有癌症中病死率高居第二名.由于机器学习方法能改进疾病预测精度,因此文章将利用它们研究肝癌前期诊断问题,提高肝癌的预测精度.首先选取影响肝癌的10个指标作为预测变量,将579位肝癌患者分为两组:随机抽取492位患者构成训练样本,剩余87位患者构成测试样本.接着利用训练样本建立6个分类器:逻辑回归、L_(2)惩罚逻辑回归、支持向量机(Support Vector Machine,SVM)、梯度提升决策树(Gradient Boosting Decision Tree,GBDT)、人工神经网络(Artificial Neural Network,ANN)和极限梯度提升算法(eXtreme Gradient Boosting,XGBoost),其中逻辑回归和L_(2)惩罚逻辑回归采用NewtonRaphson算法得到模型参数的迭代加权最小二乘估计,计算患者肿瘤细胞为恶性和良性的概率估计,确定最佳阈值预测肿瘤性状.最后用测试样本计算混淆矩阵、灵敏度和特异度,绘制ROC曲线评价预测精度.结果表明L_(2)惩罚逻辑回归预测精度最高,SVM预测精度排第二,XGBoost预测精度排第三,逻辑回归预测精度排第四,GBDT预测精度排第五,ANN和随机森林预测精度最差.

关键词： Congestive heart failure   Early detection and treatment   Telehealth monitoring   Machine learning   Exacerbation   triage   Triage  

摘要： Inadequate at-home management and self-awareness of heart failure (HF) exacerbations are known to be leading causes of the greater than 1 million estimated HF-related hospitalizations in the USA alone. Most current at-home HF management protocols include paper guidelines or exploratory health applications that lack rigor and validation at the level of the individual patient. We report on a novel triage methodology that uses machine learning predictions for real-time detection and assessment of exacerbations. Medical specialist opinions on statistically and clinically comprehensive, simulated patient cases were used to train and validate prediction algorithms. Model performance was assessed by comparison to physician panel consensus in a representative, out-of-sample validation set of 100 vignettes. Algorithm prediction accuracy and safety indicators surpassed all individual specialists in identifying consensus opinion on existence/severity of exacerbations and appropriate treatment response. The algorithms also scored the highest sensitivity, specificity, and PPV when assessing the need for emergency care. Lay summary Here we develop a machine-learning approach for providing real-time decision support to adults diagnosed with congestive heart failure. The algorithm achieves higher exacerbation and triage classification performance than any individual physician when compared to physician consensus opinion.

关键词： Machine learning   Support vector machine   Deep neural network   Geothermal heat flow   Bohai Bay Basin  

摘要： Geothermal heat flow (GHF), which contains integrated information on geothermal gradient, thermal conductivity, and heat productivity of rocks as well as crustal/mantle heat flow, is a crucial parameter in many studies including the evaluation of geothermal resources in a region and the prediction of ice sheet mass loss. The GHF measurements in the Bohai Bay Basin are insufficient and unevenly distributed, and no model has been developed to predict the GHF distribution. In this paper, using eight geological features as input data and output GHF predictions, four different machine learning algorithms were evaluated and implemented to predict the GHF data in this area: (1) a generalized linear model (GLM), (2) a deep neural network (DNN), (3) a gradient boosted regression tree (GBRT), and (4) a support vector machine (SVM). The DNN model, in which the number of neurons is a multiple of the number of features, tended to perform better than the others. Both the SVM and DNN algorithms demonstrated better accuracy, with average relative prediction errors of 13.3 and 12.7%, respectively. The accuracy improved when using a hybrid SVM and DNN approach, with an average relative prediction error of 12.2%. Based on the above models, we produced three new GHF maps of the Bohai Bay Basin. They showed similar local anomaly features but with different ranges and degrees. The new GHF maps are more detailed and rational than the map produced by kriging interpolation. The predicted GHF had much higher accuracy, which is more consistent with the measured geothermal gradient in the related areas. Finally, we reclassified the basin into four different levels of geothermal exploration target areas based on the new GHF maps.

关键词： Kernel   Magnetization   Training   Mathematical model   Principal component analysis   Magnetic domains   Computational modeling   Computational micromagnetism   deep neural networks (NNs)   kernel methods   low-rank approximation   nonlinear model order and dimensionality reduction   regularized autoencoders (AEs)  

摘要： Machine learning (ML) entered the field of computational micromagnetics only recently. The main objective of these new approaches is the automatization of solutions of parameter-dependent problems in micromagnetism, such as fast response curve estimation modeled by the Landau-Lifschitz-Gilbert (LLG) equation. Data-driven models for the solution of time- and parameter-dependent partial differential equations require high-dimensional training data structures. ML, in this case, is by no means a straightforward trivial task;it needs algorithmic and mathematical innovation. Our work introduces theoretical and computational conceptions of the certain kernel and neural network (NN)-based dimensionality reduction approaches for efficient prediction of solutions via the notion of low-dimensional feature space integration. We introduce efficient treatment of kernel ridge regression and kernel principal component analysis via low-rank approximation. The second line follows NN autoencoders as a nonlinear data-dependent dimensional reduction for the training data with a focus on accurate latent space variable description suitable for a feature space integration scheme. We verify and compare numerically by means of a NIST standard problem. The low-rank kernel method approach is fast and surprisingly accurate, while the NN scheme can even exceed this level of accuracy at the expense of significantly higher costs.

关键词： non-destructive testing   ground penetrating radar   support vector machines   ray-based method   utility pipes   parameter estimation   GprMax  

摘要： In the field of geophysics and civil engineering applications, ground penetrating radar (GPR) technology has become one of the emerging non-destructive testing (NDT) methods thanks to its ability to perform tests without damaging structures. However, NDT applications, such as concrete rebar assessments, utility network surveys or the precise localization of embedded cylindrical pipes still remain challenging. The inversion of geometric parameters, such as depth and radius of embedded cylindrical pipes, as well as the dielectric parameters of its surrounding material, is of great importance for preventive measures and quality control. Furthermore, the precise localization is mandatory for critical underground utility networks, such as gas, power and water lines. In this context, innovative signal processing techniques associated with GPR are capable of performing physical and geometric characterization tasks. This paper evaluates the performance of a supervised machine learning and ray-based methods on GPR data. Support vector machines (SVM) classification, support vector machine regression (SVR) and ray-based methods are all used to correlate information about the radius and depth of embedded pipes with the velocity of stratified media in various numerical configurations. The approach is based on the hyperbola trace emerging in a set of B-scans, given that the shape of the hyperbola varies greatly with pipe depth and radius as well as with velocity of the medium. According to the ray-based method, an inversion of the wave velocity and pipe radius is performed by applying an appropriate nonlinear least mean squares inversion technique. Feature selection within machine learning models is also implemented on the information chosen from observed hyperbola travel times. Simulated data are obtained by means of the finite-difference time-domain (FDTD) method with the 2D numerical tool GprMax. The study is carried out on mono-static, ground-coupled GPR datasets. The prelimi