Establishing the structure-property relationship in amorphous materials has been a long-term grand challenge due to the lack of a unified description of the degree of disorder.In this work,we develop SPRamNet,a neural...Establishing the structure-property relationship in amorphous materials has been a long-term grand challenge due to the lack of a unified description of the degree of disorder.In this work,we develop SPRamNet,a neural network based machine-learning pipeline that effectively predicts structure-property relationship of amorphous material via global descriptors.Applying SPRamNet on the recently discovered amorphous monolayer carbon,we successfully predict the thermal and electronic properties.More importantly,we reveal that a short range of pair correlation function can readily encode sufficiently rich information of the structure of amorphous material.Utilizing powerful machine learning architectures,the encoded information can be decoded to reconstruct macroscopic properties involving many-body and long-range interactions.Establishing this hidden relationship offers a unified description of the degree of disorder and eliminates the heavy burden of measuring atomic structure,opening a new avenue in studying amorphous materials.展开更多
Peripheral nerve injury is a serious disease and its repair is challenging. A cable-style autologous graft is the gold standard for repairing long peripheral nerve defects; however, ensuring that the minimum number of...Peripheral nerve injury is a serious disease and its repair is challenging. A cable-style autologous graft is the gold standard for repairing long peripheral nerve defects; however, ensuring that the minimum number of transplanted nerve attains maximum therapeutic effect remains poorly understood. In this study, a rat model of common peroneal nerve defect was established by resecting a 10-mm long right common peroneal nerve. Rats receiving transplantation of the common peroneal nerve in situ were designated as the in situ graft group. Ipsilateral sural nerves（10–30 mm long） were resected to establish the one sural nerve graft group, two sural nerves cable-style nerve graft group and three sural nerves cable-style nerve graft group. Each bundle of the peroneal nerve was 10 mm long. To reduce the barrier effect due to invasion by surrounding tissue and connective-tissue overgrowth between neural stumps, small gap sleeve suture was used in both proximal and distal terminals to allow repair of the injured common peroneal nerve. At three months postoperatively, recovery of nerve function and morphology was observed using osmium tetroxide staining and functional detection. The results showed that the number of regenerated nerve fibers, common peroneal nerve function index, motor nerve conduction velocity, recovery of myodynamia, and wet weight ratios of tibialis anterior muscle were not significantly different among the one sural nerve graft group, two sural nerves cable-style nerve graft group, and three sural nerves cable-style nerve graft group. These data suggest that the repair effect achieved using one sural nerve graft with a lower number of nerve fibers is the same as that achieved using the two sural nerves cable-style nerve graft and three sural nerves cable-style nerve graft. This indicates that according to the ‘multiple amplification＇ phenomenon, one small nerve graft can provide a good therapeutic effect for a large peripheral nerve defect.展开更多
Software defect prediction plays a very important role in software quality assurance,which aims to inspect as many potentially defect-prone software modules as possible.However,the performance of the prediction model ...Software defect prediction plays a very important role in software quality assurance,which aims to inspect as many potentially defect-prone software modules as possible.However,the performance of the prediction model is susceptible to high dimensionality of the dataset that contains irrelevant and redundant features.In addition,software metrics for software defect prediction are almost entirely traditional features compared to the deep semantic feature representation from deep learning techniques.To address these two issues,we propose the following two solutions in this paper:(1)We leverage a novel non-linear manifold learning method-SOINN Landmark Isomap(SL-Isomap)to extract the representative features by selecting automatically the reasonable number and position of landmarks,which can reveal the complex intrinsic structure hidden behind the defect data.(2)We propose a novel defect prediction model named DLDD based on hybrid deep learning techniques,which leverages denoising autoencoder to learn true input features that are not contaminated by noise,and utilizes deep neural network to learn the abstract deep semantic features.We combine the squared error loss function of denoising autoencoder with the cross entropy loss function of deep neural network to achieve the best prediction performance by adjusting a hyperparameter.We compare the SL-Isomap with seven state-of-the-art feature extraction methods and compare the DLDD model with six baseline models across 20 open source software projects.The experimental results verify that the superiority of SL-Isomap and DLDD on four evaluation indicators.展开更多
The development of defect prediction plays a significant role in improving software quality. Such predictions are used to identify defective modules before the testing and to minimize the time and cost. The software w...The development of defect prediction plays a significant role in improving software quality. Such predictions are used to identify defective modules before the testing and to minimize the time and cost. The software with defects negatively impacts operational costs and finally affects customer satisfaction. Numerous approaches exist to predict software defects. However, the timely and accurate software bugs are the major challenging issues. To improve the timely and accurate software defect prediction, a novel technique called Nonparametric Statistical feature scaled QuAdratic regressive convolution Deep nEural Network (SQADEN) is introduced. The proposed SQADEN technique mainly includes two major processes namely metric or feature selection and classification. First, the SQADEN uses the nonparametric statistical Torgerson–Gower scaling technique for identifying the relevant software metrics by measuring the similarity using the dice coefficient. The feature selection process is used to minimize the time complexity of software fault prediction. With the selected metrics, software fault perdition with the help of the Quadratic Censored regressive convolution deep neural network-based classification. The deep learning classifier analyzes the training and testing samples using the contingency correlation coefficient. The softstep activation function is used to provide the final fault prediction results. To minimize the error, the Nelder–Mead method is applied to solve non-linear least-squares problems. Finally, accurate classification results with a minimum error are obtained at the output layer. Experimental evaluation is carried out with different quantitative metrics such as accuracy, precision, recall, F-measure, and time complexity. The analyzed results demonstrate the superior performance of our proposed SQADEN technique with maximum accuracy, sensitivity and specificity by 3%, 3%, 2% and 3% and minimum time and space by 13% and 15% when compared with the two state-of-the-art methods.展开更多
Objective:This paper is to investigate functional connectivity differences between the abacus experts and control groups by using nonlinear processing method spatiotemporal Lyapunov exponent.Methods:11 right-handed he...Objective:This paper is to investigate functional connectivity differences between the abacus experts and control groups by using nonlinear processing method spatiotemporal Lyapunov exponent.Methods:11 right-handed healthy control children and 12 abacus children were undergone functional magnetic resonance imaging(fMRI).After preprocessing fMRI data with SPM,linear and nonlinear methods for connectivity analysis were both employed.Results:Connectivity differences between the two groups were statistically P<0.05 by the correlation method,while the P value by the nonlinear method were P<0.01.Conclusion:There are significant differences between the two groups in functional connectivity of bilateral occipital lobes.The nonlinear method proposed here seems to be more specific compared with the common linear correlation method.展开更多
针对钢材表面缺陷检测中小目标特征提取难度高、精度不足以及错检与漏检问题,文中提出一种基于改进YOLOv5s(You Only Look Once version5s)模型的小目标检测算法。通过在主干网络中引入SE(Squeeze-and-Excitation)注意力机制增强对小目...针对钢材表面缺陷检测中小目标特征提取难度高、精度不足以及错检与漏检问题,文中提出一种基于改进YOLOv5s(You Only Look Once version5s)模型的小目标检测算法。通过在主干网络中引入SE(Squeeze-and-Excitation)注意力机制增强对小目标特征的关注度。采用动态蛇形卷积(Dynamic Snake Convolution,DSConv)替换主干网络中的部分C3模块,有效提升了微弱特征的提取能力。通过采用归一化Wasserstein距离(Normalized Wasserstein Distance,NWD)优化的EIoU(Efficient Intersection over Union)损失函数降低了对小目标位置偏差的敏感性,提高了小目标的检测性能。引入解耦头优化模型头部,解决了分类与回归任务间的冲突,从而减少了错检和漏检情况的发生,提升了小目标的分类和定位准确性。在NEU-DET(Northeastern University Detection)数据集上的实验验证了所提算法的有效性,其平均精度均值(mean Average Precision,mAP)为80.4%,较原始算法提升了5%,且保持61.72 frame·s^(-1)的检测速度。结果表明,改进算法在检测速度和精度方面均优于其他对比算法,证明了其在高效检测小目标钢材表面缺陷方面的优越性。展开更多
基金supported by the National Key R&D Program of China under Grant No.2021YFA1400500the Strategic Priority Research Program of the Chinese Academy of Sciences under Grant No.XDB33000000+1 种基金the National Natural Science Foundation of China under Grant No.12334003the Beijing Municipal Natural Science Foundation under Grant Nos.JQ22001 and QY23014。
文摘Establishing the structure-property relationship in amorphous materials has been a long-term grand challenge due to the lack of a unified description of the degree of disorder.In this work,we develop SPRamNet,a neural network based machine-learning pipeline that effectively predicts structure-property relationship of amorphous material via global descriptors.Applying SPRamNet on the recently discovered amorphous monolayer carbon,we successfully predict the thermal and electronic properties.More importantly,we reveal that a short range of pair correlation function can readily encode sufficiently rich information of the structure of amorphous material.Utilizing powerful machine learning architectures,the encoded information can be decoded to reconstruct macroscopic properties involving many-body and long-range interactions.Establishing this hidden relationship offers a unified description of the degree of disorder and eliminates the heavy burden of measuring atomic structure,opening a new avenue in studying amorphous materials.
基金supported by the National Basic Research Program of China(973 Program),No.2014CB542200a grant from the Ministry of Education Innovation Team,No.IRT1201+2 种基金the National Natural Science Foundation of China,No.31271284,31171150,81171146,30971526,31100860,31040043,31640045,31671246a grant from the Educational Ministry New Century Excellent Talents Support Project in China,No.BMU20110270a grant from the National Key Research and Development Program in China,No.2016YFC1101604
文摘Peripheral nerve injury is a serious disease and its repair is challenging. A cable-style autologous graft is the gold standard for repairing long peripheral nerve defects; however, ensuring that the minimum number of transplanted nerve attains maximum therapeutic effect remains poorly understood. In this study, a rat model of common peroneal nerve defect was established by resecting a 10-mm long right common peroneal nerve. Rats receiving transplantation of the common peroneal nerve in situ were designated as the in situ graft group. Ipsilateral sural nerves（10–30 mm long） were resected to establish the one sural nerve graft group, two sural nerves cable-style nerve graft group and three sural nerves cable-style nerve graft group. Each bundle of the peroneal nerve was 10 mm long. To reduce the barrier effect due to invasion by surrounding tissue and connective-tissue overgrowth between neural stumps, small gap sleeve suture was used in both proximal and distal terminals to allow repair of the injured common peroneal nerve. At three months postoperatively, recovery of nerve function and morphology was observed using osmium tetroxide staining and functional detection. The results showed that the number of regenerated nerve fibers, common peroneal nerve function index, motor nerve conduction velocity, recovery of myodynamia, and wet weight ratios of tibialis anterior muscle were not significantly different among the one sural nerve graft group, two sural nerves cable-style nerve graft group, and three sural nerves cable-style nerve graft group. These data suggest that the repair effect achieved using one sural nerve graft with a lower number of nerve fibers is the same as that achieved using the two sural nerves cable-style nerve graft and three sural nerves cable-style nerve graft. This indicates that according to the ‘multiple amplification＇ phenomenon, one small nerve graft can provide a good therapeutic effect for a large peripheral nerve defect.
基金This work is supported in part by the National Science Foundation of China(Grant Nos.61672392,61373038)in part by the National Key Research and Development Program of China(Grant No.2016YFC1202204).
文摘Software defect prediction plays a very important role in software quality assurance,which aims to inspect as many potentially defect-prone software modules as possible.However,the performance of the prediction model is susceptible to high dimensionality of the dataset that contains irrelevant and redundant features.In addition,software metrics for software defect prediction are almost entirely traditional features compared to the deep semantic feature representation from deep learning techniques.To address these two issues,we propose the following two solutions in this paper:(1)We leverage a novel non-linear manifold learning method-SOINN Landmark Isomap(SL-Isomap)to extract the representative features by selecting automatically the reasonable number and position of landmarks,which can reveal the complex intrinsic structure hidden behind the defect data.(2)We propose a novel defect prediction model named DLDD based on hybrid deep learning techniques,which leverages denoising autoencoder to learn true input features that are not contaminated by noise,and utilizes deep neural network to learn the abstract deep semantic features.We combine the squared error loss function of denoising autoencoder with the cross entropy loss function of deep neural network to achieve the best prediction performance by adjusting a hyperparameter.We compare the SL-Isomap with seven state-of-the-art feature extraction methods and compare the DLDD model with six baseline models across 20 open source software projects.The experimental results verify that the superiority of SL-Isomap and DLDD on four evaluation indicators.
文摘The development of defect prediction plays a significant role in improving software quality. Such predictions are used to identify defective modules before the testing and to minimize the time and cost. The software with defects negatively impacts operational costs and finally affects customer satisfaction. Numerous approaches exist to predict software defects. However, the timely and accurate software bugs are the major challenging issues. To improve the timely and accurate software defect prediction, a novel technique called Nonparametric Statistical feature scaled QuAdratic regressive convolution Deep nEural Network (SQADEN) is introduced. The proposed SQADEN technique mainly includes two major processes namely metric or feature selection and classification. First, the SQADEN uses the nonparametric statistical Torgerson–Gower scaling technique for identifying the relevant software metrics by measuring the similarity using the dice coefficient. The feature selection process is used to minimize the time complexity of software fault prediction. With the selected metrics, software fault perdition with the help of the Quadratic Censored regressive convolution deep neural network-based classification. The deep learning classifier analyzes the training and testing samples using the contingency correlation coefficient. The softstep activation function is used to provide the final fault prediction results. To minimize the error, the Nelder–Mead method is applied to solve non-linear least-squares problems. Finally, accurate classification results with a minimum error are obtained at the output layer. Experimental evaluation is carried out with different quantitative metrics such as accuracy, precision, recall, F-measure, and time complexity. The analyzed results demonstrate the superior performance of our proposed SQADEN technique with maximum accuracy, sensitivity and specificity by 3%, 3%, 2% and 3% and minimum time and space by 13% and 15% when compared with the two state-of-the-art methods.
文摘Objective:This paper is to investigate functional connectivity differences between the abacus experts and control groups by using nonlinear processing method spatiotemporal Lyapunov exponent.Methods:11 right-handed healthy control children and 12 abacus children were undergone functional magnetic resonance imaging(fMRI).After preprocessing fMRI data with SPM,linear and nonlinear methods for connectivity analysis were both employed.Results:Connectivity differences between the two groups were statistically P<0.05 by the correlation method,while the P value by the nonlinear method were P<0.01.Conclusion:There are significant differences between the two groups in functional connectivity of bilateral occipital lobes.The nonlinear method proposed here seems to be more specific compared with the common linear correlation method.
文摘针对钢材表面缺陷检测中小目标特征提取难度高、精度不足以及错检与漏检问题,文中提出一种基于改进YOLOv5s(You Only Look Once version5s)模型的小目标检测算法。通过在主干网络中引入SE(Squeeze-and-Excitation)注意力机制增强对小目标特征的关注度。采用动态蛇形卷积(Dynamic Snake Convolution,DSConv)替换主干网络中的部分C3模块,有效提升了微弱特征的提取能力。通过采用归一化Wasserstein距离(Normalized Wasserstein Distance,NWD)优化的EIoU(Efficient Intersection over Union)损失函数降低了对小目标位置偏差的敏感性,提高了小目标的检测性能。引入解耦头优化模型头部,解决了分类与回归任务间的冲突,从而减少了错检和漏检情况的发生,提升了小目标的分类和定位准确性。在NEU-DET(Northeastern University Detection)数据集上的实验验证了所提算法的有效性,其平均精度均值(mean Average Precision,mAP)为80.4%,较原始算法提升了5%,且保持61.72 frame·s^(-1)的检测速度。结果表明,改进算法在检测速度和精度方面均优于其他对比算法,证明了其在高效检测小目标钢材表面缺陷方面的优越性。
