Reviewing the Variability-Overuse Injury Hypothesis: Does Movement Variability Relate to Landing Injuries?

Purpose: Overuse injuries are common in sport, but complete understanding of injury risk factors remains incomplete. Although biomechanical studies frequently examine musculoskeletal injury mechanisms, human movement variability studies aim to better understand neuromotor functioning, with proposed connections between overuse injury mechanisms and changes in motor variability. Method: In a narrative review, we discuss the variability-overuse injury hypothesis, which suggests repeated load application leads to mechanical tissue breakdown and subsequent injury when exceeding the rate of physiological adaptation. Due to the multidisciplinary nature of this hypothesis, we incorporate concepts from motor control, neurophysiology, biomechanics, as well as research design and data analysis. We therefore summarize multiple perspectives while proposing theoretical relationships between movement variability and lower extremity overuse injuries. Results: Experimental data are presented and summarized from published experiments examining interactions between experimental task demands and movement variability in the context of drop landing movements, along with comparisons to previous movement variability studies. Conclusion: We provide a conceptual framework for sports medicine researchers interested in predicting and preventing sports injuries. Under performance conditions with greater task demands, we predict reduced trial-to-trial movement variability that could increase the likelihood of overuse injuries.     

Performative technologies and teacher subjectivities: A conceptual framework

Critical educational literature suggests that an increased reliance upon performative technologies is currently transforming the very foundations from which teacher subjectivities are constructed. Arguably though, the number of studies pointing to this risk or tendency is considerably larger than the ones theorising why this should be the case. Further, in those cases where the relationship between performative technologies and teacher subjectivities is theorised, the psychological mechanisms that the technologies appeal to are seldom brought to the fore. Based on this, the purpose of this article is to theorise the psychological mechanisms that performative technologies appeal to and work through, by means of identifying, systematising and elaborating extant understandings of such mechanisms in the critical educational literature. The results are presented in the form of a conceptual framework (referred to as the CMIS-framework) which suggests that one and the same performative technology may play many different roles, where each such role appeals to and works through a particular psychological mechanism. Importantly, depending on the type of psychological mechanism that is appealed to, the CMIS-framework suggests that this will lead to teachers (un)consciously conducting particular forms of subjectivising work upon themselves, here referred to as compliance, mirroring, identification and self-realisation (CMIS).     

一种基于图卷积自编码模型的多维度学科知识网络融合方法

[目的/意义] 针对包含单一类型知识单元的知识网络难以全面反映学科知识结构的问题，提出一种从多维度进行知识网络结构融合的方法，为学科领域知识结构挖掘提供借鉴。[方法/过程] 利用LDA及TF-IDF方法抽取学科知识单元，然后运用语义相似度和关键词共现分析方法构建3个学科知识子网络：主题网络、关键词网络和实体网络，并采用空间节点传递对齐方法对齐子网络节点，接着设计基于图卷积操作的自编码模型对知识节点进行表示，最后通过计算余弦相似度重构学科知识网络。[结果/结论] 实验部分以人工智能领域为例，构建融合主题、关键词和实体的学科知识网络并展开分析，实验结果表明，本文所提方法能有效地揭示学科领域研究内容和知识结构，为学科知识发现与组织研究提供有益参考。     

小波分析技术在故障诊断中的应用

小波分析是20世纪80年代中期发展起来的新的数学理论和方法,并已在实践中得到了广泛应用.本文对小波分析、小波分析与神经网络相结合的小波网络及小波分析与模式识别相结合在故障诊断中的应用进行了详细的介绍,最后指出了这一领域有待进一步研究的问题和发展趋势.     

复合模糊神经网络在故障诊断中的应用研究

对于复杂的诊断对象,本文提出了一种复合模糊神经网络结构,该神经网络结构集成了一系列模糊神经子网络,来完成故障分类任务.文中以某高温硝酸冷却系统为诊断对象,以系统故障树为设置子网的基础,建立了含有多个子网的复合模糊神经网络.仿真结果表明了这种方法的可行性和有效性.     

警体运动损伤预防机制的研究

警察体育运动包含大量的技战术训练内容，运动损伤的类型特点与其他运动损伤差异较大。对其深入探讨有助于提高教学和训练的成效。从警察体育运动的项目特点分析出发，探索性讨论在警察体育教学和训练中运动损伤的一般性规律，提出运动损伤的具体预防机制，供广大警体教官参考。     

基于神经网络的概念联想和概念聚类

本文针对信息检索中自动标引和用户的检索提问问题 ,介绍了一种概念空间技术以及采用Hopfield神经网络进行概念联想与概念聚类的算法 ,并对其中部分算法进行了模拟和实验 ,针对存在的问题提出了改进方法 ,并通过数据说明其可行性     

基于无监督图神经网络的学术文献表示学习研究

学术文献特征表示,是学术文献搜索、分类组织、个性化推荐等学术大数据服务的关键步骤。研究表明,图神经网络能够有效学习文献的特征表示,然而当前研究主要集中在有监督学习方法上,不仅对数据集的大小和质量的要求较高,且学习到的文献特征表示与具体任务高度耦合。基于此,本文将四种无监督图神经网络方法引入学术文献表示学习,从Cora、CiteSeer和DBLP (database systems and logic programming)数据集的引文网络、共被引网络和文献耦合网络中学习文献的表示向量,并应用于文献分类和论文推荐两大下游任务。研究结果表明,(1)深度互信息图神经网络适合于文献分类任务,对抗正则化变分图自编码器则在论文推荐任务上性能更佳;(2)Cora数据集上的结果表明,相较于共被引和文献耦合网络,引文网络更适合于学习通用的文献表示向量。     

基于图像分割和EM算法的PolSAR地物分类方法

在极化合成孔径雷达（PolSAR）地物分类研究中,基于卷积神经网络的图像分割算法存在高维特征信息冗余而导致的分类边界模糊、分类精度低、计算复杂等不足,提出一种基于卷积神经网络和EM算法的轻量化图像分割网络,称为低秩重构网络（low-rank-reconstruction-net,LRR-Net）,应用于全极化SAR图像的地物分类。LRR-Net从极化目标分解的思想出发,利用EM算法对特征进行低秩重构,将特征从高维空间映射到低维空间,在减少参数的同时实现更精确的分类。用高分三号全极化图像数据对模型进行训练测试并评估,结果表明模型在保证分类精度的前提下,降低了模型复杂度。     

基于CPN网络的Web正文抽取技术研究

 通过研究使用CPN神经网络进行页面正文抽取,针对传统抽取技术在通用性、可扩展性和可维护性方面的不足提出一种解决问题的思路。