关键词： Machine learning   Genetic algorithm   Worker disability   Occupational accident   Construction safety   Safety management   Tree-based ensemble models   Artificial intelligence  

摘要： The construction industry is among the riskiest industries around the world. Hence, the preliminary studies exploring the consequences of occupational accidents have received considerable attention in research society. This study aims to develop a comprehensive framework to predict the post-accident disability status of construction workers. The dataset comprising 47,938 construction accidents recorded in Turkey was subjected to a detailed multi-step feature engineering approach, including data encoding, data scaling, dimension reduction, and data resampling. Predictions were performed through four tree-based ensemble machine learning models: Random Forest, XGBoost, AdaBoost, and Extra Trees, as well as a state-of-the-art optimization method for hyperparameter tuning, Genetic Algorithm (GA). GA-XGBoost presented the highest prediction rate with 0.8292 in terms of accuracy and 0.8120 with respect to AUROC. The findings may aid in predicting construction workers' post-accident disability status, resulting in a safer working environment and productivity planning in construction projects.

关键词： Remote sensing   Spectroscopy   Satellite   Unmanned Aerial Vehicles   Partial least squares regression   Random forests   Multiple linear regression  

摘要： The study of grasslands using machine learning (ML) methods combined with proximal/remote sensing data (RS) has been steadily increasing in the last decades. Available algorithms range from a primarily academic use to more widespread practical applications intended at helping farm management. Here, we review the use of ML methods applied to aboveground biomass (AGB) estimation in grassland systems. Based on 26 recent papers, we perform a literature review of the topic to identify common practices, namely the relation between estimation performance and the ML method used, data sources, and scale (local/regional). In order to identify the relation between the characteristics of the studies and the estimation accuracy, we use descriptive and correlation analysis. In spite of a surge in the number of papers and application examples, there is no evidence that the estimation performance of the algorithms has been improving over time. In all approaches used by the authors of the papers herein considered, the number of field samples, RS data source, and species composition of the grassland systems are the most relevant variables to explain the estimation accuracy. This accuracy increases with the number of field samples until it plateaus, hinting at the existence of an optimum level for monitoring efforts. Accuracy also increases with the proximity of the sensor to the field, i.e., on average accuracy is higher using field spectroscopy than using satellite data. There is no evidence that any particular ML method is more suited to this problem. The literature also displays significant limitations in terms of its applications of the ML algorithms. For example, a limited number of papers validated the models, casting doubt on the potential of the models for generalized application. Despite those limitations, and considering the advancements verified, we expect that, in the near future, ML methods combined with RS/proximal data will continue to improve and be helpful for farm m

关键词： classification   depression   healthcare   machine learning   prediction   random-forest  

摘要： The health sector collects a very large amount of data, hence the diagnostic process processes a very large and varied amount of data type which makes the process of analyzing these data very complicated, specifically the healthcare sector, mental health is very composed and varied by various data criteria. However, the forecast of health in modern life becomes very important. To this end, the proposed work aims to analyze patient data based on their represented symptoms, in order to help clinicians and mental health practitioners classify and refine the type of depression disorder "characterized" in patients intelligently, in order to make a relevant decision. In this context, the proposed system called CP-DDC is based on machine learning algorithms supervised more precisely by the random-forest algorithm. The dataset used in the case study contains 150 instances and 11 attributes, which define the different patient criteria, obtained from the Mohammed VI University Hospital Center of Marrakech "CHU." The results of the experiment show that the proposed system offers the highest performance.

关键词： Machine learning   Absolute information   Relative information   Ordinal classification  

摘要： The performance of an ordinal classifier is highly affected by the amount of absolute information (labelled data) available for training. In order to make up for a lack of sufficient absolute information, an effective way out is to consider additional types of information. In this work, we focus on ordinal classification problems that are provided with additional relative information. We augment several classical machine learning methods by considering both absolute and relative information as constraints in the corresponding optimization problems. We compare these augmented methods on popular benchmark datasets. The experimental results show the effectivenesses of these methods for combining absolute and relative information. (C) 2021 Elsevier B.V. All rights reserved.

关键词： Channel head extraction   Topographic attribute   Unsupervised machine learning   Clustering analysis  

摘要： Channel head extraction is fundamental to understand the catchment hydrological processes, catchment origin, runoff generation and landscape evolution. A spatially constant threshold value such as upslope area, slope-area, and curvature have been widely used for channel head extraction because of their simplicity and efficiency. However, it is very difficult to determine the threshold values within and between catchments with different topography, soil types, and land cover. In this study, a fuzzy unsupervised machine learning method (e.g., the fuzzy c means clustering) is introduced to determine the channel head locations to avoid the need for a threshold parameter. The topographic attributes (e.g., upslope area, slope, curvature, and elevation) and slope-area combined attributes (e.g., AS(2) and S/ln(A)) are used as input variables. The sensitivity of surface terrain resolution is also analyzed in terms of the proposed method's ability to predict channel heads. Two catchments with field mapped channel heads, namely Indian Creek and Mid Bailey Run in Ohio, are selected for investigation and comparison. The proposed method performs well in terms of locating the channel head. The accuracy of identified channel heads from the proposed method is comparable to published state-of-the-art channel extraction methods. Meanwhile, the proposed method has a low computational burden. Our findings also reveal the comprehensive impact of topographic attributes in locating channel heads. (C) 2021 Elsevier B.V. All rights reserved.

关键词： Neutron detection   Event reconstruction   Monte-Carlo simulations   Machine learning  

摘要： NeuLAND, the New Large Area Neutron Detector, is a key component to investigate the origin of matter in the universe with experimental nuclear physics. It is a core component of the Reactions with Relativistic Radioactive Beams setup at the Facility for Antiproton and Ion Research, Germany. Neutrons emitted from these reactions create a wide range of patterns in NeuLAND. From these patterns, the number of neutrons (multiplicity) and their first interaction points must be reconstructed to determine the neutrons' four momenta. In this paper, we detail the challenges involved in this reconstruction and present a range of possible solutions. Scikit-Learn classification models and simple Keras-based neural networks were trained on a wide range of input-scaler combinations and compared to classical models. While the improvement in multiplicity reconstruction is limited due to the overlap between features, the machine learning methods achieve a significantly better first interaction point selection, which directly improves the resolution of physical quantities.

关键词： 轨道不平顺   机器学习   Prophet模型   神经网络   时间序列预测  

摘要： 针对日常运营中火车轨道在列车荷载冲击作用下导致的轨道不平顺问题,使用机器学习方法Prophet算法和基于卷积神经网络的时间卷积网络,对轨道质量指数(TQI)数据进行建模分析和预测。对北京–上海某区段的TQI数据进行分析,并与传统模型ARIMA和三次指数平滑模型进行比较,发现其精确度更高,拟合效果更好。说明了所使用方法处理轨道不平顺数据的有效性。

关键词： 高斯过程回归   智能数值模拟框架   流致振动   圆柱  

摘要： 圆柱在横流中的流致振动响应是一个复杂的非线性过程,受到多个参数的影响。传统研究方法采用在多维度参数空间内对所有参数组合进行顺序试验的方法,研究其振动响应和受力特征,无论对于物理模型实验还是数值模拟,完成一个多参数系统的特性识别需要的工况数较多。该文采用基于高斯过程回归的机器学习方法,将人工智能与数值模拟方法相结合,建立了智能数值模拟框架。通过对已有结果进行高斯过程回归,将不确定性最大的点作为下一次数值模拟的工况,迭代直到达到收敛准则为止。通过对单参数和双参数单、双圆柱流致振动的振动响应进行预测,发现:该方法通过少量试验,即可合理预测,达到大量顺序试验的效果,显著提高了圆柱流致振动的研究效率。

关键词： Remaining useful life   Lithium-ion battery   Neural network   Long short-term memory   Early prediction  

摘要： Early prediction of battery remaining useful life (RUL) is critical to ensure a steady energy supply and the effective usage of energy. To reduce the amount of degradation data required for accurate RUL prediction, this study proposed a deep-learning-based prediction algorithm, namely ADLSTM-MC, which combines the features of adaptive dropout long short-term memory (ADLSTM) and Monte Carlo (MC) simulation. The adaptive dropout method was optimized by the Bayesian optimization. MC simulation is combined to describe the uncertainties of prediction results. To evaluate the performance of the proposed method, we developed two data sets including LiFePO4/graphite and LiNi0.8Co0.15Al0.05O2/graphite batteries with different charge/discharge rates and discharge cut-off voltages. The ADLSTM-MC method is shown to precisely achieve early prediction with only 25% of the entire degradation data, while existing models typically require 40%-70% of the degradation data. The R2 of the proposed capacity prediction model is up to 0.957-0.982, and the prediction errors including root mean square error (RMSE) and mean absolute error (MAE) are less than 0.033 and 0.027, respectively. The results show that the proposed method outperforms the mainstream regression algorithms and several recent published hybrid methods in terms of data requirement and prediction accuracy.