关键词： Climate data   Food security   Machine learning   Satellite data   Wheat yield  

摘要： Wheat is one of the most important cereal crops in the world, and its demand is expected to increase about 60% by 2050. Thus, appropriate and reliable yield forecasts are fundamental to ensure price stability and food security around the globe. In this study, we developed a Machine Learning (ML) approach to combine satellite and climate data with antecedent wheat yield information (YieldBaseLine) from 2004 - 2018, at municipal level, in Mexico. We compared the performance of four linear (generalized linear model -glm-, ridge regression -ridge-, lasso, partial least squares -pls-) and four non-linear algorithms (k-nearest neighbours -kknn-, support vector machine radial -svmR-, extreme gradient boosting -xgbTree- and random forest -rf) before harvest time. Additionally, we evaluated their performance using five different feature selection scenarios (No FS, FS = 0.9, FS = 0.75, FS = 0.9 and YieldBaseLine). The models were independently tested using two different approaches: random sampling and selective sampling. In the random sampling, the non-linear models performed generally better under the FS = 0.5 scenario, whereas the non-linear models were less sensitive to feature reduction. The results also evidenced the capacity of the YieldBaseLine predictor, combined with satellite and climate data, to address the inter-annual and spatial variability in the study area. The highest prediction accuracy was obtained by the rf method (No FS) with R-2 = 0.84. To further prove the model's operability in a simulated real-case scenario, we held out the last year records (2018) to test the models. The best performing model was again the rf (R-2 = 0.81). This study proposes a robust methodology to model crop yield (at large scale) and it may be used with operative purposes. Therefore, it can be of interest to decision and law makers, producers, authorities or the wheat industry. In addition, it can help to establish appropriate food security and trading policies. A similar approach

关键词： IDS   KNN   SVM   DT   Machine learning   Cyber security  

摘要： The increase in internet usage brings security problems with it. Malicious software can affect the operation of the systems and disrupt data confidentiality due to the security gaps in the systems. Intrusion Detection Systems (IDS) have been developed to detect and report attacks. In order to develop IDS systems, artificial intelligence based approaches have been used more frequently. In this study, literature studies using CSE-CIC IDS-2018, UNSW-NB15, ISCX-2012, NSL-KDD and CIDDS-001 data sets, which are widely used to develop IDS systems, are reviewed in detail. In addition, max-min normalization was performed on these data sets and classification was made with support vector machine (SVM), K-Nearest neighbor (KNN), Decision Tree (DT) algorithms, which are among the classical machine learning approaches. As a result, more successful results have been obtained in some of the studies given in the literature. The study is thought to be useful for developing IDS systems on the basis of artificial intelligence with approaches such as machine learning.

关键词： Vulnerable plaque   OCT   IVUS   Morphological index   Patient-specific model   FSI  

摘要： Background Coronary plaque vulnerability prediction is difficult because plaque vulnerability is non-trivial to quantify, clinically available medical image modality is not enough to quantify thin cap thickness, prediction methods with high accuracies still need to be developed, and gold-standard data to validate vulnerability prediction are often not available. Patient follow-up intravascular ultrasound (IVUS), optical coherence tomography (OCT) and angiography data were acquired to construct 3D fluid-structure interaction (FSI) coronary models and four machine-learning methods were compared to identify optimal method to predict future plaque vulnerability. Methods Baseline and 10-month follow-up in vivo IVUS and OCT coronary plaque data were acquired from two arteries of one patient using IRB approved protocols with informed consent obtained. IVUS and OCT-based FSI models were constructed to obtain plaque wall stress/strain and wall shear stress. Forty-five slices were selected as machine learning sample database for vulnerability prediction study. Thirteen key morphological factors from IVUS and OCT images and biomechanical factors from FSI model were extracted from 45 slices at baseline for analysis. Lipid percentage index (LPI), cap thickness index (CTI) and morphological plaque vulnerability index (MPVI) were quantified to measure plaque vulnerability. Four machine learning methods (least square support vector machine, discriminant analysis, random forest and ensemble learning) were employed to predict the changes of three indices using all combinations of 13 factors. A standard fivefold cross-validation procedure was used to evaluate prediction results. Results For LPI change prediction using support vector machine, wall thickness was the optimal single-factor predictor with area under curve (AUC) 0.883 and the AUC of optimal combinational-factor predictor achieved 0.963. For CTI change prediction using discriminant analysis, minimum cap thickness was the opt

关键词： Machine learning   Specific heat   Polymers  

摘要： Specific heat at constant pressure (Cp) of polymers is an important physical quantity that varies with temperature and is an essential parameter in characterizing thermal diffusivity of materials. Despite the need to predict both thermal evolution in structural composites and the resulting in situ and ex situ impacts on performance, there remain significant challenges in predicting Cp. In this work, we took an initial step toward predicting Cp at room temperature by applying machine learning (ML) methods. Decision Tree and Principal Component Analysis methods were employed in a comprehensive ML investigation. Despite the relatively small data set, the results indicate our approach is useful in designing novel polymers with desired Cp values. Our investigation about the underlying impact of polymer descriptors on Cp could pave the way for experimental syntheses of novel materials with tailored properties.

关键词： 急性缺血性脑卒中   急性卒中   治疗时间窗   临床结果   病情评估   卒中类型   CT评分   机器学习方法  

摘要： 急性缺血性脑卒中(acute ischemic stroke, AIS)是国人最常见的卒中类型,同时也是严重长期残疾的首要原因[1],由于AIS患者治疗时间窗窄,及时的病情评估和快速的诊断、治疗对于患者的预后至关重要。然而,根据Alberta卒中操作早期急性卒中分级CT评分(ASPECT)和磁共振灌注加权成像(PWI)-扩散加权成像(DWI)不匹配用作及时干预的简单指标还不能准确地预测AIS患者的临床结果[1~3]。

关键词： 泄漏区域识别   二分迭代法   k-means   随机森林分类器   特征选择   平均准确率减少  

摘要： 当采用模式识别方法识别供水管网的泄漏时,如果将每一个节点作为分类器的一个标签,由于供水管网节点泄漏特征的相似性,模型训练的准确率会比较低,因此可以通过聚类泄漏特征相似的节点形成区域,以每个区域作为分类器的标签从而提高模型训练的准确率。提出了一种基于随机森林分类器的二分迭代法识别泄漏区域,根据上一级分类器识别的泄漏区域的节点泄漏变化矩阵,采用k-means聚类将上一级迭代识别的泄漏区域聚类为两类(包含泄漏节点的区域和不包含泄漏节点的区域),从而识别包含泄漏节点的区域。随着候选泄漏区域的缩小,对识别泄漏区域有帮助的测点数量也逐渐减少,因此采用平均准确率减少(MDA)进行分类器特征(所需测点)的选择,在保证识别准确率不变的情况下减少分类器训练时所需的特征。与直接进行区域分块的识别方法相比,二分迭代法降低了选择区域分块数目时的盲目性,对于泄漏区域的识别更有目的性,提高了泄漏区域识别的准确率和效率。

关键词： 交通拥堵   拥堵判别   传统交通流理论方法   机器学习方法  

摘要： 交通拥堵是诸多交通问题中出现最多、范围最广、影响最大的问题,解决此问题的关键是对交通拥堵的判别与分析。文章从传统交通流理论与机器学习两方面综述了交通拥堵判别的方法,前者通过指标、宏观基本图、元胞自动机、元胞传输以及双流模型等模型,运用物理学与数学的理论描述交通行为特性,模型合理简单、有明确的物理意义,但限制条件较多;后者通过概率图模型、机器学习模型,建立的模型实用准确、不受固定结构约束。文中结合具体的模型方法探讨了不同拥堵判别方法的研究思路、解决方案以及存在问题,对现有的传统交通流理论方法和机器学习方法进行总结,并指出未来的发展方向。

关键词： WRF   CAMx   机器学习   PM_(2.5)   O_(3)  

摘要： 为提高西安市ρ(PM_(2.5))及ρ(O_(3))预报准确率,更好地服务西安市预报预警工作,以CAMx模式预报结果为基础,结合中尺度WRF气象预报数据、ρ(PM_(2.5))及ρ(O_(3))观测数据,基于多元线性回归、岭回归、lasso回归、决策树、随机森林以及支持向量机6种机器学习优化模型,对西安市2019年PM_(2.5)及O_(3)模拟结果进行优化.结果表明:①CAMx模式对污染物的预报存在偏差,优化模型明显修正了CAMx模式的系统性偏差,提高了预报精度.②ρ(PM_(2.5))及ρ(O_(3))的均方根误差(RMSE)由174.00、37.11μg/m^(3)分别降至34.36~39.37、24.77~28.82μg/m^(3),相关性系数(R)由0.63、0.78分别提至0.70~0.78、0.83~0.88.③不同模型对模拟值的订正优势不同,随机森林对PM_(2.5)优化效果显著,优化提高率为80%;支持向量机对O_(3)的优化效果最理想,优化提高率为36%;线性回归方法对O_(3)的优化效果较好,但对PM_(2.5)的优化效果相对较差.研究显示,机器学习模型显著优化了CAMx模拟结果,反映了利用机器学习修正空气质量数值模式预报结果的研究意义和可行性.

关键词： Fault detection   CWRU bearing fault database   Performance criteria   Classification   Pattern recognition   Machine learning  

摘要： This paper presents a systematic procedure to fairly compare experimental performance scores for machine learning methods for fault diagnosis based on vibration signals. In the vast majority of related scientific publications, the estimated accuracy and similar performance criteria are the sole quality parameter presented. However, the experimental design giving rise to these results is mostly biased, based on unacceptably simple validation methods and on recycling identical patterns in test data sets, previously used for training. Moreover, the methods in general overfit their hyperparameters, introducing additional overoptimistic results. In order to remedy this defect, we critically analyse the usual training-validation-test division and propose an algorithmic guideline in the form of a validation framework. This allows a well defined comparison of experimental results. In order to illustrate the ideas of the paper, the Case Western Reserve University Bearing Data benchmark is used as a case study. Four distinct classifiers are experimentally compared, under gradually more difficult generalization tasks using the proposed evaluation framework: K-Nearest-Neighbor, Support Vector Machine, Random Forest and One-Dimensional Convolutional Neural Network. An extensive literature review suggests that most vibration based research papers, particularly for the Case Western Reserve University Bearing Data, use similar patterns for training and testing, making their classification an easy task.

关键词： Lane-changing recognition   Convolutional neural network   Lane-changing behaviour   Freeway off-ramps  

摘要： Crashes are occurred frequently at freeway off-ramps due to improper lane-changing (LC) behaviours. The LC behaviour is the main cause of freeway off-ramp crashes. It is important to warn the LC to reduce potential crashes. The uncertainty of LC behaviour increases the difficulties of predicting in advance. The off-ramps at Xi'an Raocheng freeway were chosen for investigation. The datasets were collected by the UAV. There was a total of 637 LC images extracted from the 200 minutes' video stream. All LC behaviours were divided into twelve categories according to the changing direction and the influence of other vehicles in the target-lane or ego-lane. The machine learning technology is efficient in the image recognition. Thus, the vision technology was applied to devised a lane-changing recognition (LCR) model with the two-level convolutional neural network. A novel convolutional neural network based on the AlexNet was also proposed to compare with the LCR model. All samples were divided into a training dataset and a testing dataset for two models. The performance of two machines networks was compared. The training average accuracy was above 94.6% with the LCR model. The LCR model outperformed the model based on the AlexNet which was only 73.97% on average. (C) 2020 Elsevier B.V. All rights reserved.