基于大数据挖掘技术的地震舆情感知研究

对地震舆情信息的深入感知和有效管理,能够保障社会和谐发展.提出一个基于大数据技术和深度学习的地震舆情感知平台,基于Hadoop和MongoDB大数据技术实现对海量实时地震舆情数据的处理和存储.基于Word2vec和LSTM的融合模型能够有效实现震后网民的情感识别,为舆情预警提供支持.以台湾5.8级地震舆情数据为例,对该...     

提高GPS定位精度的群调度滤波算法*

为解决现行安全生产科技领域中全球定位系统（GPS）/惯性导航系统（INS）融合系统容易产生较大时变误差的关键问题,满足GPS定位信号的精度和信号数据处理时间的要求,提出1种群调度滤波算法,通过实验仿真,分别与模糊算法、机器学习算法及卡尔曼滤波网络算法进行比较研究。结果表明:群调度滤波算法兼顾了定位精度与数据处理时间,具有较大优势,可在航空安全、船舶安全、应急监测、灾后抢险等安全科技领域广泛应用。     

Development of machine learning based prediction models for hazardous properties of chemical mixtures

Lower flammability limit (LFL), upper flammability limit (UFL), auto-ignition temperature (AIT) and flash point (FP) are crucial hazardous properties for fire and explosion hazards assessment and consequence analysis. In this study, a comprehensive prediction model set was constructed by using expanded chemical mixture databases of chemical mixture hazardous properties. Machine learning based gradient boosting quantitative structure-property relationship (GB-QSPR) method is implemented for the first time to improve the model performance and prediction accuracy. The result shows that all developed models have significantly higher accuracy than other regular QSPR models, with the 5-fold cross-validation RMSE of LFL, UFL, AIT, and FP models being 1.06, 1.14, 1.08, and 1.17, respectively. All developed QSPR models can be used to estimate reliable chemical mixture hazardous properties and provide useful guidance in chemical mixture hazard assessment and consequence analysis.     

面向实时人群动力学分析的深度基本图网络

当前行人疏散实验中基本图计算方法通常是通过对每个行人进行跟踪实现的.但这种跟踪方法难以实现实时人群动力学分析.针对这一问题,提出了深度基本图网络.实验提出的网络框架由两个模块组成,即多尺度递归卷积神经网络(MSR-Net)和光流模块,分别对行人密度和行人速度进行估计.具体来讲,MSR-Net学习了输入图像与行人密度图之...     

Machine learning for radioxenon event classification for the Comprehensive Nuclear-Test-Ban Treaty

A method of weapon detection for the Comprehensive nuclear-Test-Ban-Treaty (CTBT) consists of monitoring the amount of radioxenon in the atmosphere by measuring and sampling the activity concentration of ^131mXe, ¹³³Xe, ^133mXe, and ¹³⁵Xe by radionuclide monitoring. Several explosion samples were simulated based on real data since the measured data of this type is quite rare. These data sets consisted of different circumstances of a nuclear explosion, and are used as training data sets to establish an effective classification model employing state-of-the-art technologies in machine learning. A study was conducted involving classic induction algorithms in machine learning including Naïve Bayes, Neural Networks, Decision Trees, k-Nearest Neighbors, and Support Vector Machines, that revealed that they can successfully be used in this practical application. In particular, our studies show that many induction algorithms in machine learning outperform a simple linear discriminator when a signal is found in a high radioxenon background environment.     

The importance of context for effective public engagement: learning from the governance of waste

This paper takes a new look at the importance of context – institutional and political – in effective public engagement processes. It does so through a rare comparative opportunity to examine the effectiveness of processes of public engagement in two UK waste authorities, where the same waste company was involved as both the primary contractor for the delivery of the waste management service (including new energy-from-waste facilities) and, furthermore, the same staff delivered the public engagement. Interrogating these cases affords the opportunity to place flesh on the bones of the sometimes ‘abstract’ skeleton of context. While engagement processes support effective local governance in an era of partnerships and deliberative democracy, the paper identifies that the methods adopted cannot be played out devoid of detailed understanding and response to local context, including the strength of partnership working between the public and private sector, the degree of political support for engagement, and the extent to which a traditional institutional paternalism still dominates.     

Social learning in a policy-mandated collaboration: community wildfire protection planning in the eastern United States

Policies such as the US Healthy Forests Restoration Act (HFRA) mandate collaboration in planning to create benefits such as social learning and shared understanding among partners. However, some question the ability of top-down policy to foster successful local collaboration. Through in-depth interviews and document analysis, this paper investigates social learning and transformative learning in three case studies of Community Wildfire Protection Planning (CWPP), a policy-mandated collaboration under HFRA. Not all CWPP groups engaged in social learning. Those that did learned most about organisational priorities and values through communicative learning. Few participants gained new skills or knowledge through instrumental learning. CWPP groups had to commit to learning, but the design of the collaborative-mandate influenced the type of learning that was most likely to occur. This research suggests a potential role for top-down policy in setting the structural context for learning at the local level, but also confirms the importance of collaborative context and process in fostering social learning.     

Social Learning Through Evaluation: A Case Study of Overcoming Constraints for Management of Dryland Salinity

Conventional approaches to evaluation of environmental programs have tended to limit themselves to restricted measures of program effectiveness. This paper shows how a social learning approach can be incorporated into evaluating public environmental programs. A social learning approach is particularly suited to complex environmental challenges which are inherently difficult to understand, predict, and manage, thus complicating the evaluation process. The paper presents an Australian case study of dryland salinity management where there are major knowledge barriers impeding conventional management techniques. The research presented in this paper focused on evaluating a public demonstration program to track its impact through its design, implementation, and monitoring phases. The paper shows that, by incorporating social learning principles and practices, program evaluation can promote collective action, critical reflection, and increased knowledge to underpin improved environmental management.     

基于模糊神经网络的机动车驾驶教练员职业适宜性研究

提出应用模糊神经网络系统,建构教练员职业适宜性的3个要素(即心理素质、驾驶技能和知识的表达阐述能力)的学习样本,分析"三要素"的8项特征参数指标——场依存性、速度估计能力、交通安全意识、简单反应、选择反应、跟踪能力、行车注意力、表达阐述能力等,使用K-均值法对实验样本进行初始分类,形成标准学习样本,使用该样本对所构建系统进行训练和调试。利用经调试训练后的系统,依据所测教练员的心理、心理参数对其职业适宜性进行评价。试验表明:建立的教练员职业适宜性仿真模型能取得很好的评价效果。     

浅谈高校办公室工作的基本规律与工作技巧

从明晰高校办公室工作职责、高校办公室主任应具备的能力、创新高校办公室管理模式三个方面,阐述了高校办公室工作的基本规律。通过在高校办公室多年的工作积累,总结出了高校办公室主任八方面的工作技巧。