检索结果-南通市图书馆

同方期刊数据库

Federated Meta Reinforcement Learning for Personalized Tasks

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2024年第3期29卷 911-926页

作者： wentao liu Xiaolong Xu Jintao Wu Jielin Jiang School of Computer Science Nanjing University of Information Science and TechnologyNanjing 210044China School of Software Nanjing University of Information Science and TechnologyNanjing 210044China

As an emerging privacy-preservation machine learning framework,Federated Learning(FL)facilitates different clients to train a shared model collaboratively through exchanging and aggregating model parameters while raw data are kept local and *** this learning framework is applied to Deep Reinforcement Learning(DRL),the resultant Federated Reinforcement Learning(FRL)can circumvent the heavy data sampling required in conventional DRL and benefit from diversified training data,besides privacy preservation offered by *** FRL implementations presuppose that clients have compatible tasks which a single global model can *** practice,however,clients usually have incompatible(different but still similar)personalized tasks,which we called task *** may severely hinder the implementation of FRL for practical *** this paper,we propose a Federated Meta Reinforcement Learning(FMRL)framework by integrating Model-Agnostic Meta-Learning(MAML)and ***,we innovatively utilize Proximal Policy Optimization(PPO)to fulfil multi-step local training with a single round of ***,considering the sensitivity of learning rate selection in FRL,we reconstruct the aggregation optimizer with the Federated version of Adam(Fed-Adam)on the server *** experiments demonstrate that,in different environments,FMRL outperforms other FL methods with high training efficiency brought by Fed-Adam.

关键词： federated learning reinforcement learning meta-learning personalization

B cell development and antibody responses in human immune system mice:current status and future perspective

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Life Sciences) 2024年第4期67卷 645-652页

作者： Tao Zhang wentao liu Yong-Guang Yang Key Laboratory of Organ Regeneration&Transplantation of the Ministry of Education The First Hospital of Jilin UniversityChangchun 130061China National-Local Joint Engineering Laboratory of Animal Models for Human Diseases Jilin UniversityChangchun 130061China International Center of Future Science Jilin UniversityChangchun 130061China

Humanized immune system(HIS)mice have been developed and used as a small surrogate model to study human immune function under normal or disease *** variations are found between models,most HIS mice show robust human T cell ***,there has been unsuccessful in constructing HIS mice that produce high-affinity human antibodies,primarily due to defects in terminal B cell differentiation,antibody affinity maturation,and development of primary follicles and germinal *** this review,we elaborate on the current knowledge about and previous attempts to improve human B cell development in HIS mice,and propose a potential strategy for constructing HIS mice with improved humoral immunity by transplantation of human follicular dendritic cells(FDCs)to facilitate the development of secondary follicles.

关键词： humanized immune system mouse humoral immunity germinal center B cell follicular dendritic cell

同方期刊数据库

Quasinormal modes and isospectrality of Bardeen(Anti-)de Sitter black holes

在线全文

学校读者我要写书评

暂无评论

Chinese Physics C 2024年第3期48卷 191-202页

作者：赵莹刘文韬张超房熊俊荆继良 Department of Physics Key Laboratory of Low Dimensional Quantum Structures and Quantum Control of Ministry of Educationand Synergetic Innovation Center for Quantum Effects and ApplicationsInstitute of Interdisciplinary StudiesHunan Normal UniversityChangsha 410081China Department of Physics Faculty of ScienceTokyo University of Science1-3KagurazakaShinjuku-kuTokyo 162-8601Japan

Black holes(BHs)exhibiting coordinate singularities but lacking essential singularities throughout the spacetime are referred to as regular black holes(RBHs).The initial formulation of RBHs was presented by Bardeen,who considered the Einstein equation coupled with a nonlinear electromagnetic *** this study,we investigate the gravitational perturbations,including the axial and polar sectors,of the Bardeen(Anti-)de Sitter black *** derive the master equations with source terms for both axial and polar perturbations and subsequently compute the quasinormal modes(QNMs)through numerical *** the Bardeen de Sitter black hole,we employ the 6thorder WKB *** numerical results reveal that the isospectrality is broken in this ***,the QNM frequencies are calculated using the HH method for the Bardeen Anti-de Sitter black hole.

关键词： Bardeen quasinormal modes master equation

在线全文

学校读者我要写书评

暂无评论

论违法减资情形下股东对债权人的责任

河南大学学报（社会科学版） 2024年第1期64卷 43-49,153页

作者：刘文涛郑州西亚斯学院文法学院河南郑州451100

我国现行《公司法》安排的减资制度,本意是回应债权人利益保护,但是仅有程序规则,欠缺责任规则,尚不足以完全实现减资制度初衷。股东因违法减资对债权人承担的责任,与类推适用的抽逃出资责任、瑕疵出资责任,因应不同的行为及责任基础,... 详细信息

我国现行《公司法》安排的减资制度,本意是回应债权人利益保护,但是仅有程序规则,欠缺责任规则,尚不足以完全实现减资制度初衷。股东因违法减资对债权人承担的责任,与类推适用的抽逃出资责任、瑕疵出资责任,因应不同的行为及责任基础,系属相互独立的责任类型,应当为其设定独立的请求权基础规范。股东对债权人承担补充赔偿责任足以衡平违法减资对法益的损坏,不逾越股东有限责任的边界,也非放纵股东背信行为。从司法成本与债权实现效率的角度观察,也应当赋予债权人在违法减资情形下对股东的直接求偿权。

关键词：违法减资债权人保护股东责任

Identification of cell surface markers for acute myeloid leukemia prognosis based on multi-model analysis

在线全文

学校读者我要写书评

暂无评论

Journal of Biomedical Research 2024年第4期38卷 397-412页

作者： Jiaqi Tang Lin Luo Bakwatanisa Bosco Ning Li Bin Huang Rongrong Wu Zihan Lin Ming Hong Wenjie liu Lingxiang Wu Wei Wu Mengyan Zhu Quanzhong liu Peng Xia Miao Yu Diru Yao Sali Lv Ruohan Zhang wentao liu Qianghu Wang Kening Li Department of Bioinformatics School of Biomedical Engineering and InformaticsNanjing Medical UniversityNanjingJiangsu 211166China Department of Hematology of the Affiliated Huai'an No.1 People's Hospital of Nanjing Medical University Northern Jiangsu Institute of Clinical MedicineHuai'anJiangsu 223300China Collaborative Innovation Center for Personalized Cancer Medicine Jiangsu Key Lab of Cancer BiomarkersPrevention and TreatmentNanjing Medical UniversityNanjingJiangsu 211166China Department of Hematology the First Affiliated Hospital of Nanjing Medical UniversityJiangsu Province HospitalNanjingJiangsu 210029China Key Laboratory of Hematology of Nanjing Medical University NanjingJiangsu 210029China Department of Pharmacology School of Basic Medical SciencesNanjing Medical UniversityNanjingJiangsu 211166China The Affiliated Cancer Hospital of Nanjing Medical University Jiangsu Cancer HospitalJiangsu Institute of Cancer ResearchNanjingJiangsu 210002China

Given the extremely high inter-patient heterogeneity of acute myeloid leukemia(AML),the identification of biomarkers for prognostic assessment and therapeutic guidance is *** surface markers(CSMs)have been shown to play an important role in AML leukemogenesis and *** the current study,we evaluated the prognostic potential of all human CSMs in 130 AML patients from The Cancer Genome Atlas(TCGA)based on differential gene expression analysis and univariable Cox proportional hazards regression *** using multi-model analysis,including Adaptive LASSO regression,LASSO regression,and Elastic Net,we constructed a 9-CSMs prognostic model for risk stratification of the AML *** predictive value of the 9-CSMs risk score was further validated at the transcriptome and proteome *** Cox regression analysis showed that the risk score was an independent prognostic factor for the AML *** AML patients with high 9-CSMs risk scores had a shorter overall and event-free survival time than those with low ***,single-cell RNA-sequencing analysis indicated that patients with high 9-CSMs risk scores exhibited chemotherapy ***,PI3K inhibitors were identified as potential treatments for these high-risk *** conclusion,we constructed a 9-CSMs prognostic model that served as an independent prognostic factor for the survival of AML patients and held the potential for guiding drug therapy.

关键词： acute myeloid leukemia cell surface markers prognosis drug sensitivity multi-model analysis

改进型DeepLabV3+的糖尿病眼底病变分割

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

华中科技大学学报（自然科学版） 2024年第5期52卷 90-97页

作者：马晓普刘文涛李贺南阳师范学院计算机科学与技术学院河南南阳473061 南阳师范学院生命科学与农业工程学院河南南阳473061

针对糖尿病视网膜眼底病变多类分割难及小病灶识别率低的问题,提出一种融合注意力机制与改进型DeepLabV3+的糖尿病视网膜眼底病变多类病症分割方法.该方法首先通过使用MobileNetV2网络提取病灶特征,减少参数量并提高算法训练速度;接着... 详细信息

针对糖尿病视网膜眼底病变多类分割难及小病灶识别率低的问题,提出一种融合注意力机制与改进型DeepLabV3+的糖尿病视网膜眼底病变多类病症分割方法.该方法首先通过使用MobileNetV2网络提取病灶特征,减少参数量并提高算法训练速度;接着通过优化空洞空间卷积金字塔池化中的空洞卷积层数与空洞率,以提高捕获小病灶特征的能力;然后基于DeepLabV3+网络模型进行改进,借助坐标注意力机制感知病灶方向和位置信息,从而提高识别精度;最后采用FGADR和IDRiD数据集分别训练和测试所提出的模型.实验结果表明:所提出方法的平均交并比(MIoU)指标为73.75%,具有较高的分割精度,验证了模型有效性.

关键词：糖尿病视网膜眼底病变深度学习 DeepLabV3+网络坐标注意力多类分割

同方期刊数据库

杂萘联苯聚芳醚酮表面水凝胶涂层的制备及其抗菌性能

在线全文

学校读者我要写书评

暂无评论

合成树脂及塑料 2024年第1期41卷 1-7页

作者：柳承德张阿利刘文韬蹇锡高大连理工大学化工学院辽宁大连116024 大连理工大学辽宁省高性能树脂工程技术研究中心辽宁大连116024

在杂萘联苯聚芳醚酮(PPEK)与银粉掺杂制备的复合材料表面造孔后,采用银离子催化表面自由基聚合法制备聚丙烯酸/丙烯酰胺水凝胶涂层。通过傅里叶变换红外光谱仪、钨灯丝扫描电子显微镜、X射线光电子能谱仪等对水凝胶涂层的结构与性能进... 详细信息

在杂萘联苯聚芳醚酮(PPEK)与银粉掺杂制备的复合材料表面造孔后,采用银离子催化表面自由基聚合法制备聚丙烯酸/丙烯酰胺水凝胶涂层。通过傅里叶变换红外光谱仪、钨灯丝扫描电子显微镜、X射线光电子能谱仪等对水凝胶涂层的结构与性能进行表征,并研究水凝胶涂层改性PPEK的抑菌作用及其细胞相容性。结果表明:PPEK表面造孔后,涂层与基体材料的结合能从185.83 J/m^(2)提高到926.26 J/m^(2);该水凝胶涂层对大肠杆菌和金黄色葡萄球菌的抑菌率达到95%以上,且无细胞毒性,扩大了PPEK在医疗卫生领域的应用。

关键词：杂萘联苯聚芳醚酮水凝胶涂层自由基聚合抗菌活性细胞相容性

高速跨介质入水多相流动与流固耦合特性研究综述

同方期刊数据库博看期刊评论

在线全文

学校读者我要写书评

暂无评论

空气动力学学报 2024年第1期42卷 67-85,I0002页

作者：明付仁王嘉捷刘文韬刘祥聚张阿漫哈尔滨工程大学船舶工程学院哈尔滨150001

高速跨介质入水问题广泛存在于海洋工程、航空航天等领域,入水多相流动与流固耦合作用机理是该领域的研究热点,对于跨介质航行器的载荷预报、弹道稳定性评估、结构强度校核以及安全性设计等具有重要意义。本文围绕高速跨介质入水所涉及... 详细信息

高速跨介质入水问题广泛存在于海洋工程、航空航天等领域,入水多相流动与流固耦合作用机理是该领域的研究热点,对于跨介质航行器的载荷预报、弹道稳定性评估、结构强度校核以及安全性设计等具有重要意义。本文围绕高速跨介质入水所涉及的基础关键力学问题,重点论述了高速跨介质多相流动与空泡演化、冲击载荷与降载方法、运动稳定性与流固耦合响应、流固耦合数值研究方法等方面的研究现状,并针对目前仍存在的问题提出了建议,旨在为高速跨介质入水的相关研究与设计等提供参考。

关键词：高速入水多相流动冲击载荷与降载结构响应流固耦合

基于BS-TabNet和LSSA的车架智能轻量化设计

在线全文

学校读者我要写书评

暂无评论

湖南大学学报（自然科学版） 2024年第2期51卷 163-176页

作者：聂昕刘文涛陈少伟张承霖陈勇杨昊湖南大学整车先进设计制造技术全国重点实验室湖南长沙410082 株洲中车时代电气股份有限公司湖南株洲412000 湖大艾盛汽车技术开发有限公司湖南长沙410205

为解决传统牵引车车架轻量化设计中设计周期长、设计难度较大和过分依赖工程师经验等问题,提出了一种智能轻量化设计方法.首先,通过试验设计(Design of Experiments,DOE)联合仿真获取车架性能表格数据.其次,基于TabNet算法、贝叶斯优化... 详细信息

为解决传统牵引车车架轻量化设计中设计周期长、设计难度较大和过分依赖工程师经验等问题,提出了一种智能轻量化设计方法.首先,通过试验设计(Design of Experiments,DOE)联合仿真获取车架性能表格数据.其次,基于TabNet算法、贝叶斯优化算法和沙普利增量解释理论(SHapley Additive exPlanation,SHAP)构建出BS-TabNet模型,用于学习车架性能表格数据,生成车架代理模型.最后,采用莱维飞行策略对麻雀搜索算法(Sparrow Search Algorithm,SSA)进行改进,得到莱维麻雀搜索算法(Levy Sparrow Search Algorithm,LSSA),用于求解车架轻量化任务,找到最优车架结构参数.相较于传统机器学习算法,BS-TabNet模型在准确性、稳定性和可解释性三个方面都有着更高的评价,其准确度达到了0.98左右,稳定性提高50%以上,而且具有更强的可解释能力,解决了深度学习在表格型数据上表现较差的问题.相较于传统群智能优化算法,LSSA算法能够寻找到更好的优化结果,在满足其他性能要求的同时,实现了车架质量减轻5.64%的轻量化效果.智能轻量化设计方法将人工智能与车架轻量化设计相结合,能够节省大量设计时间,提高设计效率.

关键词：深度学习贝叶斯优化车架设计 TabNet SHAP SSA

维普期刊数据库博看期刊