Results 31 to 40 of about 22,554 (129)
智能科学与技术学报, 2019 针对髋关节软骨的应力分布算法研究问题，设计了一个基于深度学习模型来代替有限元分析。该深度学习模型分为无监督学习模块和有监督学习模块，首先使用无监督学习模块对髋关节的软骨和股骨进行形状编码；之后实现对应力分布数据的编码与解码，使得应力数据能够与神经网络相结合；然后通过监督学习，利用编码好的应力数据进行监督，使神经网络学习得到一个从髋关节软骨和股骨的形状码到应力分布的应力码的映射关系；最终得到一个拟合的深度学习模型。此模型能够在一定程度上模拟有限元分析方法，但是由于其平均绝对误差和归一化平均绝对误差比较大 ...
刘远平 +3 more
doaj
Geographic risks to functional groups of mammals and birds from habitat loss in Mexico
Conservation Biology, EarlyView.Abstract There is a need to quantify the impact of habitat loss due to anthropogenic factors on different aspects of biodiversity, such as functional trait diversity represented by functional groups (FGs). We developed a metric to assess the weighted risk of loss of habitat for 33 FGs of mammals and 36 FGs of birds in Mexico based on potential ...
Fernando Mayani‐Parás +4 more
wiley +1 more source
Conservation Biology, EarlyView.Abstract Constructed value of information (CVoI) is an expert elicitation decision‐analytic tool used to prioritize sources of uncertainty based on their potential to improve decision outcomes if resolved. Despite increased application of CVoI, the robustness of CVoI prioritization of sources of uncertainty relative to differences in expert elicitation
Kristin P. Davis +7 more
wiley +1 more source
智能科学与技术学报, 2019 概述了联邦可视化的概念、框架、方法与应用。联邦可视化框架能够在不进行数据整合的情况下，针对具体任务和特定场景进行加密训练，得出反映全体数据特征的可视化模型。联邦可视化是联邦学习框架在可视化领域的拓展应用，主要强调在保障数据隐私的前提下，互利共赢的联邦协作方式在对多数据源数据进行可视分析方面的应用，以打破各领域、各行业的数据壁垒，实现数据与知识的共享。
魏雅婷, 王智勇, 周舒悦, 陈为
doaj
智能科学与技术学报, 2021 在电力系统需要巡检的大环境下，人工巡检的传统方式存在很大不便和安全隐患，而采用无人机的目标检测方法在绝缘子检测识别方向有很大的应用前景。针对绝缘子图像检测中存在的场景复杂、视角多变、设备计算能力受限等问题，提出了一种改进的轻量级SE-YOLOv5s卷积神经网络来实现对绝缘子的快速目标检测，该方法首先在YOLOv5s网络中融合SE注意力模块，以强化网络对绝缘子目标的辨识能力，随后采用K-means聚类方法构建绝缘子的先验框，以提升定位精度，最后构造置信度与定位任务联合的损失函数，并引入 Mosaic ...
田庆 +4 more
doaj
Comprehensive Risk Assessment of Hazards in Grain Supply Chain Based on Multi-Dimensional Data(多维数据驱动粮食供应链危害物风险综合评价) [PDF]
Shipin kexue jishu xuebao, 2019 In order to scientifically and reasonably evaluate the comprehensive risks of heavy metals, mycotoxins, pesticide residues and microorganisms in all aspects of the grain supply chain, based on the analysis of sampling data and other multi-dimensional ...
WANG Xiaoyi(王小艺) +5 more
doaj +1 more source
Conservation Biology, EarlyView.Abstract Achieving global conservation policy goals requires the ability to set and measure progress toward science‐based targets for biodiversity. The species threat abatement and restoration (STAR) metric was developed to enable actors to set science‐based targets for species.
Louise Mair +17 more
wiley +1 more source
智能科学与技术学报, 2019 随着智能交通的发展，快速、精确识别交通场景成为亟待解决的重要问题。目前已有许多识别方法可以提高交通场景的识别效果，但这些算法无法提取视觉概念的交通语义特征，导致识别精度低下。为此，设计了一种提取高维场景语义特征和结构信息的识别算法，以提高识别精度。为减少图像高维与低维特征表示之间的“语义鸿沟”，首先构建了一个场景类的语义描述系统，然后通过最小化损失（element-wise logistic loss）函数训练多标签分类网络，获取交通场景图像的高维特征表示，最后在4个大规模场景识别数据集上进行验证 ...
刘文华 +4 more
doaj
International Journal of Applied Linguistics, EarlyView.ABSTRACT This computational linguistic study analyzes 8743 Rednote posts to examine how sentiment, emotion, and content domain collectively shape social comparison dynamics on social media. We identify four key findings: (1) an emotional gradient, where upward comparison content exhibits the highest sentiment scores and downward comparison the lowest; (
Michelle Mingyue Gu +7 more
wiley +1 more source
Accountability in Government Use of AI: Citizen Concerns and Preferences
Public Administration, EarlyView.ABSTRACT Government adoption of AI opens the “pandora's box” of accountability, a core aspect of public value. Despite the centrality of governmental AI accountability, citizens' concerns and preferences in holding governments accountable for AI usage remain poorly understood.
Naikang Feng, Yanto Chandra
wiley +1 more source