Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss fun...Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss function or rebalancing the classifier.However,one crucial aspect overlooked by previous research studies is the imbalanced feature space problem caused by the imbalanced angle distribution.In this paper,the authors shed light on the significance of the angle distribution in achieving a balanced feature space,which is essential for improving model performance under long-tailed distributions.Nevertheless,it is challenging to effectively balance both the classifier norms and angle distribution due to problems such as the low feature norm.To tackle these challenges,the authors first thoroughly analyse the classifier and feature space by decoupling the classification logits into three key components:classifier norm(i.e.the magnitude of the classifier vector),feature norm(i.e.the magnitude of the feature vector),and cosine similarity between the classifier vector and feature vector.In this way,the authors analyse the change of each component in the training process and reveal three critical problems that should be solved,that is,the imbalanced angle distribution,the lack of feature discrimination,and the low feature norm.Drawing from this analysis,the authors propose a novel loss function that incorporates hyperspherical uniformity,additive angular margin,and feature norm regularisation.Each component of the loss function addresses a specific problem and synergistically contributes to achieving a balanced classifier and feature space.The authors conduct extensive experiments on three popular benchmark datasets including CIFAR-10/100-LT,ImageNet-LT,and iNaturalist 2018.The experimental results demonstrate that the authors’loss function outperforms several previous state-of-the-art methods in addressing the challenges posed by imbalanced and longtailed datasets,that is,by improving upon the best-performing baselines on CIFAR-100-LT by 1.34,1.41,1.41 and 1.33,respectively.展开更多
Molecular electrocatalysts have demonstrated potential for the hydrogen evolution reaction(HER)due to their well-defined structures and high intrinsic activities.Achieving rapid production of hydrogen requires molecul...Molecular electrocatalysts have demonstrated potential for the hydrogen evolution reaction(HER)due to their well-defined structures and high intrinsic activities.Achieving rapid production of hydrogen requires molecular electrocatalysts to operate at high current densities,which still presents a challenge.In this work,we demonstrate that molecularly dispersed electrocatalysts of cobalt phthalocyanine anchored on carbon nanotubes(CoPc MDEs)are superior candidates due to the efficient charge transport between the substrate and the active site.The intrinsic activity can be enhanced by introducing functional groups on phthalocyanine.To facilitate mass transport,di(ethylene glycol)substituted CoPc molecules are further anchored on a threedimensional self-supported electrode(CoPc-DEG MDE@CC),enabling continuous operation for 25 h at−1000 mA/cm^(2)in 1.0 M KOH.Our study demonstrates the potential of molecular electrocatalysts for HER and emphasizes the importance of adjusting intrinsic activity,and charge and mass transport capacity for practical molecular electrocatalysts.展开更多
It has been known that,the novel coronavirus,2019-nCoV,which is considered similar to SARS-CoV,invades human cells via the receptor angiotensin converting enzyme II(ACE2).Moreover,lung cells that have ACE2 expression ...It has been known that,the novel coronavirus,2019-nCoV,which is considered similar to SARS-CoV,invades human cells via the receptor angiotensin converting enzyme II(ACE2).Moreover,lung cells that have ACE2 expression may be the main target cells during 2019-nCoV infection.However,some patients also exhibit non-respiratory symptoms,such as kidney failure,implying that 2019-nCoV could also invade other organs.To construct a risk map of different human organs,we analyzed the single-cell RNA sequencing(scRNA-seq)datasets derived from major human physiological systems,including the respiratory,cardiovascular,digestive,and urinary systems.Through scRNA-seq data analyses,we identified the organs at risk,such as lung,heart,esophagus,kidney,bladder,and ileum,and located specific cell types(i.e.,type II alveolar cells(AT2),myocardial cells,proximal tubule cells of the kidney,ileum and esophagus epithelial cells,and bladder urothelial cells),which are vulnerable to 2019-nCoV infection.Based on the findings,we constructed a risk map indicating the vulnerability of different organs to 2019-nCoV infection.This study may provide potential clues for further investigation of the pathogenesis and route of 2019-nCoV infection.展开更多
基金National Key Research and Development Program of China,Grant/Award Numbers:2022YFB3103900,2023YFB3106504Major Key Project of PCL,Grant/Award Numbers:PCL2022A03,PCL2023A09+5 种基金Shenzhen Basic Research,Grant/Award Number:JCYJ20220531095214031Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies,Grant/Award Number:2022B1212010005Shenzhen International Science and Technology Cooperation Project,Grant/Award Number:GJHZ20220913143008015Natural Science Foundation of Guangdong Province,Grant/Award Number:2023A1515011959Shenzhen-Hong Kong Jointly Funded Project,Grant/Award Number:SGDX20230116091246007Shenzhen Science and Technology Program,Grant/Award Numbers:RCBS20221008093131089,ZDSYS20210623091809029。
文摘Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss function or rebalancing the classifier.However,one crucial aspect overlooked by previous research studies is the imbalanced feature space problem caused by the imbalanced angle distribution.In this paper,the authors shed light on the significance of the angle distribution in achieving a balanced feature space,which is essential for improving model performance under long-tailed distributions.Nevertheless,it is challenging to effectively balance both the classifier norms and angle distribution due to problems such as the low feature norm.To tackle these challenges,the authors first thoroughly analyse the classifier and feature space by decoupling the classification logits into three key components:classifier norm(i.e.the magnitude of the classifier vector),feature norm(i.e.the magnitude of the feature vector),and cosine similarity between the classifier vector and feature vector.In this way,the authors analyse the change of each component in the training process and reveal three critical problems that should be solved,that is,the imbalanced angle distribution,the lack of feature discrimination,and the low feature norm.Drawing from this analysis,the authors propose a novel loss function that incorporates hyperspherical uniformity,additive angular margin,and feature norm regularisation.Each component of the loss function addresses a specific problem and synergistically contributes to achieving a balanced classifier and feature space.The authors conduct extensive experiments on three popular benchmark datasets including CIFAR-10/100-LT,ImageNet-LT,and iNaturalist 2018.The experimental results demonstrate that the authors’loss function outperforms several previous state-of-the-art methods in addressing the challenges posed by imbalanced and longtailed datasets,that is,by improving upon the best-performing baselines on CIFAR-100-LT by 1.34,1.41,1.41 and 1.33,respectively.
基金supported by Guangdong-Hong Kong-Macao Joint Laboratory for Photonic-Thermal-Electrical Energy Materials and Devices(No.2019B121205001)Shenzhen fundamental research funding(Nos.JCYJ20220818100618039 and JCYJ20200109141405950)+1 种基金the National Natural Science Foundation of China(No.22075125)supported by the Center for Computational Science and Engineering(SUSTech).
文摘Molecular electrocatalysts have demonstrated potential for the hydrogen evolution reaction(HER)due to their well-defined structures and high intrinsic activities.Achieving rapid production of hydrogen requires molecular electrocatalysts to operate at high current densities,which still presents a challenge.In this work,we demonstrate that molecularly dispersed electrocatalysts of cobalt phthalocyanine anchored on carbon nanotubes(CoPc MDEs)are superior candidates due to the efficient charge transport between the substrate and the active site.The intrinsic activity can be enhanced by introducing functional groups on phthalocyanine.To facilitate mass transport,di(ethylene glycol)substituted CoPc molecules are further anchored on a threedimensional self-supported electrode(CoPc-DEG MDE@CC),enabling continuous operation for 25 h at−1000 mA/cm^(2)in 1.0 M KOH.Our study demonstrates the potential of molecular electrocatalysts for HER and emphasizes the importance of adjusting intrinsic activity,and charge and mass transport capacity for practical molecular electrocatalysts.
基金This work was supported in part by the China National Science and Technology Major Project for Prevention and Treatment of Infectious Diseases(No.2017ZX10203207 to Z.-G.H.)National Natural Science Foundation of China(No.81672772 to Z.-G.H.,No.31601070 to J.H.,No.31800253 to K.C.)+1 种基金Interdisciplinary Program of Shanghai Jiao Tong University(Nos.2019TPA09 and ZH2018ZDA33 to Z.-G.H.,J.H.,and X.Z.)Shanghai Sailing Program(No.17YF1410400 to K.C.)and Innovative Research Team of High-Level Local Universities in Shanghai.
文摘It has been known that,the novel coronavirus,2019-nCoV,which is considered similar to SARS-CoV,invades human cells via the receptor angiotensin converting enzyme II(ACE2).Moreover,lung cells that have ACE2 expression may be the main target cells during 2019-nCoV infection.However,some patients also exhibit non-respiratory symptoms,such as kidney failure,implying that 2019-nCoV could also invade other organs.To construct a risk map of different human organs,we analyzed the single-cell RNA sequencing(scRNA-seq)datasets derived from major human physiological systems,including the respiratory,cardiovascular,digestive,and urinary systems.Through scRNA-seq data analyses,we identified the organs at risk,such as lung,heart,esophagus,kidney,bladder,and ileum,and located specific cell types(i.e.,type II alveolar cells(AT2),myocardial cells,proximal tubule cells of the kidney,ileum and esophagus epithelial cells,and bladder urothelial cells),which are vulnerable to 2019-nCoV infection.Based on the findings,we constructed a risk map indicating the vulnerability of different organs to 2019-nCoV infection.This study may provide potential clues for further investigation of the pathogenesis and route of 2019-nCoV infection.
