关键词： River mouth width   remote sensing   statistic approach  

摘要： Prediction of river mouth morphological variation plays a crucial role in the study of unique brackish river ecosystems and is important for economic development. However, it is still a challenge to assess river mouth morphological variation because it is influenced by multiple factors. With the increasing availability of observation data, machine learning techniques have been widely used in the geosciences. This study aims to examine different machine learning techniques (logistic regression, neural network, bootstrap aggregating, AdaBoost, LogitBoost, and random subspace) for the prediction of estuary morphological variation probability. The Da Dien estuary, located in Phu Yen Province, Vietnam, was selected as a case study. To construct the machine learning models, the throat width of the Da Dien estuary (the dependent variable) was determined using Landsat 4, 5, 7, and 8. Six independent variables (wind energy, wind direction, swell energy, swell direction, tidal prism, and river discharge) were obtained from either local hydrological stations or worldwide data portals. Through the examination of several statistical indicators, such as receiver operating characteristic, area under receiving operating characteristic, root-mean-square error, mean absolute error, and kappa, the results indicate that among the examined techniques, the Random Subspace model outperformed all others in almost every indicator, and it provided the most accurate prediction of estuary throat width variation. There are some limitations to this approach, such as the requirement of a large amount of data and difficulties in putting human-induced factors and topographical variables into the model. However, this paper presents a promising approach for providing a rapid assessment and can aid in developing a conceptual model of the system.

关键词： Multi-objective optimization   Spotted hyena optimizer   Salp swarm algorithm   Cancer classification   Gene expression   Machine learning   Gene selection   Feature selection  

摘要： Background and objective: Over the last two decades, DNA microarray technology has emerged as a powerful tool for early cancer detection and prevention. It helps to provide a detailed overview of disease complex microenvironment. Moreover, online availability of thousands of gene expression assays made microarray data classification an active research area. A common goal is to find a minimum subset of genes and maximizing the classification accuracy. Methods: In pursuit of a similar objective, we have proposed framework (C-HMOSHSSA) for gene selection using multi-objective spotted hyena optimizer (MOSHO) and salp swarm algorithm (SSA). The real-life optimization problems with more than one objective usually face the challenge to maintain convergence and diversity. Salp Swarm Algorithm (SSA) maintains diversity but, suffers from the overhead of maintaining the necessary information. On the other hand, the calculation of MOSHO requires low computational efforts hence is used for maintaining the necessary information. Therefore, the proposed algorithm is a hybrid algorithm that utilizes the features of both SSA and MOSHO to facilitate its exploration and exploitation capability. Results: Four different classifiers are trained on seven high-dimensional datasets using a subset of features (genes), which are obtained after applying the proposed hybrid gene selection algorithm. The results show that the proposed technique significantly outperforms existing state-of-the-art techniques. Conclusion: It is also shown that the new sets of informative and biologically relevant genes are successfully identified by the proposed technique. The proposed approach can also be applied to other problem domains of interest which involve feature selection. (C) 2019 Elsevier B.V. All rights reserved.

摘要： Findings from these models can be used to identify those patients that are likely to develop CDI following surgical *** The study was funded by the Brodie Craniofacial *** of interest None for all *** A Supplementary data The following is the supplementary data to this article: Characteristic Odds Ratio (95% CI) p-value Age Each 1 unit increase 1.01 (1.01–1.01) <0.01 Sex Female 0.89 (0.87–0.91) <0.01 Male Reference Race African American 1.15 (1.10–1.19) <0.01 Hispanic 0.98 (0.93–1.03) 0.38 Asian/Pacific Islander 1.20 (1.09–1.31) <0.01 Native American 1.10 (0.90–1.34) 0.34 Other Races 1.02 (0.94–1.11) 0.59 Caucasian Reference Comorbid burden Each 1 unit increase 1.46 (1.45–1.47) <0.01 Insurance status Medicare 1.23 (1.19–1.28) <0.01 Medicaid 1.41 (1.34–1.48) <0.01 Uninsured 1.10 (1.01–1.20) 0.03 Other insurance 0.94 (0.86–1.03) 0.22 Private insurance Reference Location/teaching status of hospital Urban teaching hospital 1.39 (1.34–1.44) <0.01 Urban non-teaching/rural hospitals Reference Geographic region Northeast 1.20 (1.14–1.26) <0.01 Midwest 0.89 (0.84–0.94) <0.01 South 0.88 (0.83–0.92) <0.01 West Reference Year of hospitalization 2013 0.95 (0.91–1.00) 0.04 2014 0.91 (0.87–0.96) <0.01 2012 Reference Table 1 Characteristics associated with CDI during hospitalization (traditional multivariable logistic regression model). Rule set model FPR TPR Rule set 1 - Comorbid ≥ 5, Age ≥ 72, and location = northeast 0.009 0.038 Rule set 2 - Comorbid ≥ 5 and urban teaching hospital 0.085 0.270 Rule set 3 - Comorbid ≥ 5 0.141 0.435 Rule set 4 - Comorbid ≥ 3 - Race = African American, Asian/Pacific Islander, or Native American 0.491 0.813 Table 2 Bayesian Rule Set models and their performances (Machine Learning Model).

关键词： 大数据时代   正刊   信息技术   优化能力   信号模型   机器学习方法   信号特征   智能信号处理  

摘要： 随着信息技术的高速发展,在现代社会生活中每天都有着各式各样的大量数据产生。大数据时代已然来临,而如何合理地存储、处理以及利用这些数据就成为了受到广泛关注的焦点性问题。一方面,大数据更加依赖具有学习能力、推理能力和优化能力的智能信号处理方法;另一方面,大数据为智能信号处理方法的发展提供了条件。大数据时代的智能信号处理是以信号模型为驱动,以大样本为基础,通过机器学习方法提取信号特征,进而实现对信号的智能化处理和利用。本次征文的主要目的是给大数据时代下的智能信号处理提供新的视角和思路。

关键词： 机器翻译系统   机器学习方法  

摘要： 通信技术的进步对各行各业都产生了重大影响,尤其是教育领域。世界各地的人们只需要上网就可以实时听取诺贝尔奖获得者的演讲,或者在线学习最优秀大学的课程并获得相应的学分。但是,如果用户无法理解演讲者的语言,则无法从在线音视频中获得有用的信息。为了解决这个问题,日本奈良科学技术研究所(NAIST)的科学家在日本信息处理协会自然语言处理兴趣小组(IPSJ SIG-NL)第240次会议上提出一种新的基于机器学习的解决方案。

关键词： 马里兰大学   北卡罗来纳大学   人工智能   机器学习方法   情绪   教堂山   行走   性能表现  

摘要： 情绪在人们的生活中扮演着十分重要的角色,及时、准确地看出同伴或是他人的情绪,能够帮助我们更好地与他人进行交流、沟通。最近,美国北卡罗来纳大学教堂山分校和马里兰大学帕克分校的研究人员研究了一种全新的机器学习方法,这种方法可以根据个体在行走时的姿势识别出他/她的情绪。目前,这种方法在实验中取得了良好的性能表现。

关键词： 技术群组   专利文献   文本挖掘   聚类分析   趋势预测  

摘要： 以国家知识产权局专利数据库为数据来源,采用专利文本挖掘、聚类分析和技术预测相结合的方法,结合R-stideo软件和Loglet Lab软件中的统计分析工具,针对煤化工技术进行技术群组分类及趋势预测研究。研究表明:煤化工技术可分成六大技术群组,即污水处理技术群组、煤气净化技术群组、煤气化技术群组、传统煤化工群组、污染物监测技术群组以及煤液化技术群组。传统煤化工技术专利申请量已进入下降期,表明传统煤化工技术进入衰退期,未来向更加清洁的新型煤化工方向发展;煤气净化技术、煤气化技术和煤液化技术等,经过发展进入稳定发展期,市场空间巨大;污水处理技术和污染物检测技术作为污染治理方面的技术处于快速发展阶段,污水处理技术正朝着更加高效、更加环保以及方法更加综合的方向发展,煤气净化技术正向着更加经济、环保、可靠和高效的方向发展。

关键词： 扩散加权成像   前列腺癌   机器学习  

摘要： 探究使用机器学习方法,提升对扩散加权成像(DWI)多参数图的前列腺癌(PCa)诊断的准确性。对39例前列腺癌患者、56例良性患者,进行磁共振扩散加权图像的采集,并使用传统单指数模型(Mono)、拉伸指数模型(SEM)、弥散张量成像(DTI)模型、弥散峰度成像(DKI)模型以及体内素不相干运动扩散(IVIM)模型等5种重建模型,得到共计16个参数图,而后对于每一个参数图进行直方图分析,得到相关图像特征后使用机器学习的方法进行分类。使用支持向量机和随机森林两种分类器对前列腺病变进行良恶性分类,随机森林分类器的AUC值可以达到0.98,具有较高的分类性能。另外,对特征进行重要性排序后,发现DKI参数图是肿瘤分类的重要指标。

关键词： 打入桩   轴向承载力   极端梯度提升   回归算法   机器学习  

摘要： 在复杂的工程地质环境中,打入桩的轴向承载力的预测是设计和施工的一个重大挑战。本研究将利用机器学习工具一极端梯度提升算法(XGBoost)预测打入桩的轴向承载力,对桩身最大压缩应力(MCS)、最大拉伸应力(MTS)和每英尺锤击数(BPF)进行了研究,并与反向传播神经网络(BPNN)和随机森林(RF)算法进行了比较分析。利用美国北卡罗来纳州的桩数据库中4000多个数据集生成训练和预测样本。利用均方根误差(RMSE)、平均绝对误差(MAE)和线性相关系数(R^2)等性能指标对三种回归算法的有效性进行了验证。结果表明,与其他两种方法相比.XGBoost算法在解决桩、锤、土以及施工工艺等复杂非线性问题时,具有更高的稳定性和预测精度。XGBoost算法作为一种可靠的机器学习算法,可以为其他类似桩基工程轴向承载力的预测提供参考。

关键词： 机器学习   嘉兴   小微企业   信用风险   实证分析  

摘要： 以嘉兴市2013-2017年发生信贷业务的小微企业为研究对象,对2646家小微企业2013-2017年的金融信用信息进行实证分析,构建了5类小微企业信用风险评估模型,结果显示,随机森林模型更适用于嘉兴市小微企业信用风险评估,根据该方法将小微企业风险分为“高档”“中档”“低档”3类,提出了政策建议。