目的采用双聚类方法对人工智能在医学领域的国际研究成果进行分析,探讨主题领域内热点趋势。方法检索Web of Science核心合集数据库中医学人工智能的相关文献,采用Co-Occurrence13.4提取高频关键词生成词篇矩阵,应用gCluto1.0聚类工具...目的采用双聚类方法对人工智能在医学领域的国际研究成果进行分析,探讨主题领域内热点趋势。方法检索Web of Science核心合集数据库中医学人工智能的相关文献,采用Co-Occurrence13.4提取高频关键词生成词篇矩阵,应用gCluto1.0聚类工具包进行双聚类分析。结果共纳入文献7803篇,年发文量整体呈上升趋势,美国位居发文总量的首位,共提取30个高频主题词,形成人工智能应用于生物标志物检测等6个聚类。研究热点聚焦于卫生保健、疾病转归、疾病全程监测、辅助诊断癌症、预测模型效验和鉴别生物标志物6个主题。结论人工智能已普遍应用于临床诊断和治疗,为基因检测及公共卫生事件提供了针对性的支持,但国内相关研究还处于发展阶段,未来还需要依托多学科、机构间的交流合作,推动中国智能化医疗的发展,使其真正成为促进医疗卫生事业发展的重要工具。展开更多
对称IB(Symmetric Information Bottleneck)通过行、列压缩变量之间的相互协作来挖掘数据中的双向压缩模式.由于行、列压缩变量不能完全承载行、列基层变量中所蕴含的特征信息,从而导致对称IB所得的数据双向压缩模式与基层变量所蕴含的...对称IB(Symmetric Information Bottleneck)通过行、列压缩变量之间的相互协作来挖掘数据中的双向压缩模式.由于行、列压缩变量不能完全承载行、列基层变量中所蕴含的特征信息,从而导致对称IB所得的数据双向压缩模式与基层变量所蕴含的内在模式之间存在一定的偏离.针对该问题,通过最大化地保存压缩变量与基层变量交叉之间的互信息,将基层变量引入到数据的双向压缩中,使它们协助压缩变量共同来学习联合分布中的双向压缩模式,提出交叉对称IB:ICSIB(Inter-Correlated Symmetric Information Bottleneck).ICSIB算法采用交错的顺序"抽取-合并"迭代过程来优化压缩变量与基层变量交叉之间的互信息,可保证得到目标函数的一个局部优解.实验结果表明,在基层特征变量的协助下,ICSIB算法得到的数据双向压缩模式更接近于数据中真实的内在模式,并可有效地应用于数据的联合聚类中.展开更多
Objective:To make a visual analysis of literature metrology on the training of applied nursing talents in China,to clarify the research hotspots and development trends,and to provide basic information for the educatio...Objective:To make a visual analysis of literature metrology on the training of applied nursing talents in China,to clarify the research hotspots and development trends,and to provide basic information for the education and teaching reform of applied nursing talents in China.Methods:The literature related to the training of applied nursing talents in China was retrieved from CNKI,Wanfang,Weipu database.NoteExpress was used for removing duplicates,and Bicomb2.0 was used to extract key fields and generate word matrix,which was imported into gCluto1.0 for double cluster analysis.Results:Finally,675 articles met the inclusion and exclusion criteria,the number of articles is increasing yearly.39 high-frequency keywords,such as practice teaching and 9 core authors such as Song Mei are extracted.The results of double cluster analysis showed that the research topics included five aspects:nursing education reform;training of applied nursing talents at different levels;quality of practical teaching;practice teaching reform and laboratory management;application-oriented nursing personnel training mode.Conclusion:The reform and practice of application-oriented nursing talents training in our country are developing continuously,and has gradually received great attention from universities,the local governments and national government departments.The research focuses on the reform of training mode of applied nursing talents at different levels,practice teaching reform,laboratory management and other aspects,which can provide a reference for the reform and practice of applied nursing talents training in China.展开更多
文摘目的采用双聚类方法对人工智能在医学领域的国际研究成果进行分析,探讨主题领域内热点趋势。方法检索Web of Science核心合集数据库中医学人工智能的相关文献,采用Co-Occurrence13.4提取高频关键词生成词篇矩阵,应用gCluto1.0聚类工具包进行双聚类分析。结果共纳入文献7803篇,年发文量整体呈上升趋势,美国位居发文总量的首位,共提取30个高频主题词,形成人工智能应用于生物标志物检测等6个聚类。研究热点聚焦于卫生保健、疾病转归、疾病全程监测、辅助诊断癌症、预测模型效验和鉴别生物标志物6个主题。结论人工智能已普遍应用于临床诊断和治疗,为基因检测及公共卫生事件提供了针对性的支持,但国内相关研究还处于发展阶段,未来还需要依托多学科、机构间的交流合作,推动中国智能化医疗的发展,使其真正成为促进医疗卫生事业发展的重要工具。
文摘对称IB(Symmetric Information Bottleneck)通过行、列压缩变量之间的相互协作来挖掘数据中的双向压缩模式.由于行、列压缩变量不能完全承载行、列基层变量中所蕴含的特征信息,从而导致对称IB所得的数据双向压缩模式与基层变量所蕴含的内在模式之间存在一定的偏离.针对该问题,通过最大化地保存压缩变量与基层变量交叉之间的互信息,将基层变量引入到数据的双向压缩中,使它们协助压缩变量共同来学习联合分布中的双向压缩模式,提出交叉对称IB:ICSIB(Inter-Correlated Symmetric Information Bottleneck).ICSIB算法采用交错的顺序"抽取-合并"迭代过程来优化压缩变量与基层变量交叉之间的互信息,可保证得到目标函数的一个局部优解.实验结果表明,在基层特征变量的协助下,ICSIB算法得到的数据双向压缩模式更接近于数据中真实的内在模式,并可有效地应用于数据的联合聚类中.
基金This work was supported by the State Nationalities Commission(No.19077).
文摘Objective:To make a visual analysis of literature metrology on the training of applied nursing talents in China,to clarify the research hotspots and development trends,and to provide basic information for the education and teaching reform of applied nursing talents in China.Methods:The literature related to the training of applied nursing talents in China was retrieved from CNKI,Wanfang,Weipu database.NoteExpress was used for removing duplicates,and Bicomb2.0 was used to extract key fields and generate word matrix,which was imported into gCluto1.0 for double cluster analysis.Results:Finally,675 articles met the inclusion and exclusion criteria,the number of articles is increasing yearly.39 high-frequency keywords,such as practice teaching and 9 core authors such as Song Mei are extracted.The results of double cluster analysis showed that the research topics included five aspects:nursing education reform;training of applied nursing talents at different levels;quality of practical teaching;practice teaching reform and laboratory management;application-oriented nursing personnel training mode.Conclusion:The reform and practice of application-oriented nursing talents training in our country are developing continuously,and has gradually received great attention from universities,the local governments and national government departments.The research focuses on the reform of training mode of applied nursing talents at different levels,practice teaching reform,laboratory management and other aspects,which can provide a reference for the reform and practice of applied nursing talents training in China.
