Results 31 to 40 of about 5,473,456 (145)
大数据, 2023 医学命名实体识别是从医学文本中抽取出指代特定概念的医学实体，是医学信息抽取的基础性任务。当前主流的医学命名实体识别算法普遍基于深度学习技术，需要大量高质量的标注样本进行模型训练。然而医学领域的样本标注成本很高，严重限制了模型性能的提升。为了降低模型对标注样本的需求，一种重要的方法是基于主动学习思想，设计合理的样本采样策略，自动选取高价值样本优先标注，从而使模型提前收敛。现有算法普遍基于样本长度、样本识别的概率等特征来设计采样策略，忽视了样本类别分布这一深层次特征，导致命名实体识别召回率较低 ...
刘聪, 吕雪峰, 王宏林, 王晓伟, 陆瑾, 孙顺, 胡松奇
doaj +1 more source
Gaoya dianqi, 2023 为了研究电力电缆局部放电的模式识别，解决传统单一智能算法识别率低的问题，文中提出了一种融合多深度学习算法的混合智能算法。首先，设计并制作5种典型缺陷模型以模拟实际电力电缆中的缺陷，据此展开实验并收集数据；然后，通过对PRPD谱图的相窗归一化、去极端值等改进，以及绘制PRCD谱图，更全面凸显局部放电有用特征；最后，训练基于PRPD或PRCD的多种深度学习分算法，通过可信度融合得到混合智能算法。实验结果表明，该混合智能算法相比常规单一深度学习算法识别率有显著提升，总体可达98.504 ...
杨朝锋 +6 more
doaj
International Insolvency Review, Volume 35, Issue 2, Page 406-435, Summer 2026.Abstract Although a local experiment, the promulgation of the Regulations of Shenzhen Special Economic Zone on Personal Bankruptcy (SPBR) in 2020 was hailed as a significant milestone in China's insolvency lawmaking which has thus far addressed only corporate insolvencies.
Jenny Fu, Jin Chun
wiley +1 more source
Rare Metals, Volume 45, Issue 8, August 2026.ABSTRACT Advanced metallic glasses, leveraging their ultrahigh strength, excellent corrosion resistance, superior soft magnetic properties and good biocompatibility, have gradually become indispensable strategic cornerstones in critical fields such as aerospace, electronic information and energy power. This unique combination of properties enables them
Yizhou Liu +11 more
wiley +1 more source
智能科学与技术学报, 2021 随着深度学习技术的突破和大型数据集的提出，行人轨迹预测的准确度已经成为人工智能领域的研究热点之一。主要对行人轨迹预测的技术分类和研究现状进行详细的综述。根据模型建模方式的不同，将现有方法分为基于浅层学习的轨迹预测方法和基于深度学习的轨迹预测方法，分析了每类方法中具有代表性的算法的效果及优缺点，归纳了当前主流的轨迹预测公开数据集，并在数据集中对比了主流轨迹预测方法的性能，最后对轨迹预测技术面临的挑战与发展趋势进行了展望。
李琳辉, 周彬, 任威威, 连静
doaj
Multimodal Deep Learning‐Assisted Microstructural Evaluation System for Sintered NdFeB Magnets
Rare Metals, Volume 45, Issue 8, August 2026.ABSTRACT The performance of sintered NdFeB magnets, especially coercivity, is critically determined by their complex microstructure. However, a quantitative microstructural evaluation system (MES) is much needed but has long been missing in the community.
Zihao Wang +8 more
wiley +1 more source
, 2019 针对电力负荷预测中存在的随机性、不确定性的问题,结合深度学习算法具有很强的自适应感知能力等特点,采用目前较为主流的深度学习方法,如长短时记忆(LSTM)网络、门循环单元(GRU)神经网络和栈式自编码器(SAE),分别研究其应用于电力负荷预测时的效果。研究发现,将历史负荷数据作为三种深度学习预测模型的输入时,三种预测模型的负荷预测精度指标评估结果各有不同。因此,为了全面评估三种预测模型的预测效果,提出将不同时间段内的相同历史负荷数据作为预测模型输入对比各模型的负荷预测精度,从中找出最佳的预测模型 ...
张建寰, 吉莹, 陈立东
core +1 more source
Structured machine learning modeling to support conservation of deep‐sea benthic biodiversity
Conservation Biology, Volume 40, Issue 4, August 2026.Abstract Biodiversity monitoring programs need to deliver accurate, timely, and actionable predictions. To establish a predictive monitoring program for deep‐sea benthos of the Santos Basin, Brazil, we developed a two‐stage structured model that allowed comparison of biodiversity predictions obtained from environmental simulations (2M‐Sim).
Gustavo Fonseca +23 more
wiley +1 more source
He jishu随着高能核物理研究进入多维度、高复杂度数据分析阶段，深度学习技术正逐步成为理解极端条件下核物质行为的关键工具，并推动研究范式从经验驱动向数据驱动的根本转变。本文简要梳理了机器学习在该领域的演进，并着重介绍了深度学习方法在其中的前沿进展：早期（20世纪末至21世纪10年代）研究主要采用人工神经网络和支持向量机等传统算法，通过核质量预测、相变识别等任务验证了机器学习处理核物理问题的可行性，但受限于人工特征提取和计算能力的制约，尚未触及物理特征的自主挖掘；深度学习时代（21世纪10年代至今 ...
张 靖宗 +4 more
doaj +1 more source
International Journal of Applied Linguistics, Volume 36, Issue 3, Page 2098-2109, August 2026.ABSTRACT This study conducted linguistic analyses of the written corrective feedback (WCF) for Chinese as a second language (CSL) provided by chat generative pre‐trained transformer (ChatGPT) and human teachers (including preservice teachers and senior teachers).
Ling Zhang +2 more
wiley +1 more source