This study proposed a weighted sampling hierarchical classification learning method based on an efficient backbone network model to address the problems of high costs,low accuracy,and time-consuming traditional tea di...This study proposed a weighted sampling hierarchical classification learning method based on an efficient backbone network model to address the problems of high costs,low accuracy,and time-consuming traditional tea disease recognition methods.This method enhances the feature extraction ability by conducting hierarchical classification learning based on the EfficientNet model,effectively alleviating the impact of high similarity between tea diseases on the model’s classification performance.To better solve the problem of few and unevenly distributed tea disease samples,this study introduced a weighted sampling scheme to optimize data processing,which not only alleviates the overfitting effect caused by too few sample data but also balances the probability of extracting imbalanced classification data.The experimental results show that the proposed method was significant in identifying both healthy tea leaves and four common leaf diseases of tea(tea algal spot disease,tea white spot disease,tea anthracnose disease,and tea leaf blight disease).After applying the“weighted sampling hierarchical classification learning method”to train 7 different efficient backbone networks,most of their accuracies have improved.The EfficientNet-B1 model proposed in this study achieved an accuracy rate of 99.21%after adopting this learning method,which is higher than EfficientNet-b2(98.82%)and MobileNet-V3(98.43%).In addition,to better apply the results of identifying tea diseases,this study developed a mini-program that operates on WeChat.Users can quickly obtain accurate identification results and corresponding disease descriptions and prevention methods through simple operations.This intelligent tool for identifying tea diseases can serve as an auxiliary tool for farmers,consumers,and related scientific researchers and has certain practical value.展开更多
Good distribution of samples and weights can improve the computational accuracy and efficiency in the stochastic response analyses of aerospace problems with uncertain parameters.This work proposes a new Generalized L...Good distribution of samples and weights can improve the computational accuracy and efficiency in the stochastic response analyses of aerospace problems with uncertain parameters.This work proposes a new Generalized L2 Discrepancy based on a General Point(GL2D-GP)for generating samples and their corresponding weights.The proposed GL2D-GP is an extension of the existing discrepancy by introducing the non-same weights and a smaller box to measure probability errors.Minimizing the GL2D-GP can yield a weight optimization formula that generates a set of optimal non-identical weights for a given sample set.Through minimizing the GL2D-GP assigned to the set of optimal non-same weights,a new sample and weight generation method is developed.In the proposed method,the samples can be easily generated in terms of the generalized Halton formula with a series of optimal permutation vectors which are found by the intelligent evolutionary algorithm.Once the sample set is obtained,the optimal weights can be generated in terms of the weight optimization formula.Five numerical examples are presented to verify the high accuracy,efficiency,and strong robustness of the proposed sample generation method based on GL2D-GP.展开更多
Probabilistic assessment of seismic performance(SPPA)is a crucial aspect of evaluating the seismic behavior of structures.For complex bridges with inherent uncertainties,conducting precise and efficient seismic reliab...Probabilistic assessment of seismic performance(SPPA)is a crucial aspect of evaluating the seismic behavior of structures.For complex bridges with inherent uncertainties,conducting precise and efficient seismic reliability analysis remains a significant challenge.To address this issue,the current study introduces a sample-unequal weight fractional moment assessment method,which is based on an improved correlation-reduced Latin hypercube sampling(ICLHS)technique.This method integrates the benefits of important sampling techniques with interpolator quadrature formulas to enhance the accuracy of estimating the extreme value distribution(EVD)for the seismic response of complex nonlinear structures subjected to non-stationary ground motions.Additionally,the core theoretical approaches employed in seismic reliability analysis(SRA)are elaborated,such as dimension reduction for simulating non-stationary random ground motions and a fractional-maximum entropy single-loop solution strategy.The effectiveness of this proposed method is validated through a three-story nonlinear shear frame structure.Furthermore,a comprehensive reliability analysis of a real-world long-span,single-pylon suspension bridge is conducted using the developed theoretical framework within the OpenSees platform,leading to key insights and conclusions.展开更多
目的利用自适应合成抽样(adaptive synthetic sampling,ADASYN)与类别逆比例加权法处理类别不平衡数据,结合分类器构建模型对阿尔茨海默病(alzheimer′s disease,AD)患者疾病进程进行分类预测。方法数据源自阿尔茨海默病神经影像学计划(...目的利用自适应合成抽样(adaptive synthetic sampling,ADASYN)与类别逆比例加权法处理类别不平衡数据,结合分类器构建模型对阿尔茨海默病(alzheimer′s disease,AD)患者疾病进程进行分类预测。方法数据源自阿尔茨海默病神经影像学计划(Alzheimer′s disease neuroimaging initiative,ADNI),经随机森林填补缺失值,弹性网络筛选特征子集后,利用ADASYN与类别逆比例加权法处理类别不平衡数据。分别结合随机森林(random forest,RF)、支持向量机(support vector machine,SVM)构建四种模型:ADASYN-RF、ADASYN-SVM、加权随机森林(weighted random forest,WRF)、加权支持向量机(weighted support vector machine,WSVM),与RF、SVM比较分类性能。模型评价指标为宏观平均精确率(macro-average of precision,macro-P)、宏观平均召回率(macro-average of recall,macro-R)、宏观平均F1值(macro-average of F1-score,macro-F1)、准确率(accuracy,ACC)、Kappa值和AUC(area under the ROC curve)。结果ADASYN-RF的分类性能最优(Kappa值为0.938,AUC为0.980),ADASYN-SVM次之。利用ADASYN-RF预测得到的重要分类特征分别为CDRSB、LDELTOTAL、MMSE,在临床上均可得到证实。结论ADASYN与类别逆比例加权法都能辅助提升分类器性能,但ADASYN算法更优。展开更多
基金financial support provided by the Major Project of Yunnan Science and Technology,under Project No.202302AE09002003,entitled“Research on the Integration of Key Technologies in Smart Agriculture.”。
文摘This study proposed a weighted sampling hierarchical classification learning method based on an efficient backbone network model to address the problems of high costs,low accuracy,and time-consuming traditional tea disease recognition methods.This method enhances the feature extraction ability by conducting hierarchical classification learning based on the EfficientNet model,effectively alleviating the impact of high similarity between tea diseases on the model’s classification performance.To better solve the problem of few and unevenly distributed tea disease samples,this study introduced a weighted sampling scheme to optimize data processing,which not only alleviates the overfitting effect caused by too few sample data but also balances the probability of extracting imbalanced classification data.The experimental results show that the proposed method was significant in identifying both healthy tea leaves and four common leaf diseases of tea(tea algal spot disease,tea white spot disease,tea anthracnose disease,and tea leaf blight disease).After applying the“weighted sampling hierarchical classification learning method”to train 7 different efficient backbone networks,most of their accuracies have improved.The EfficientNet-B1 model proposed in this study achieved an accuracy rate of 99.21%after adopting this learning method,which is higher than EfficientNet-b2(98.82%)and MobileNet-V3(98.43%).In addition,to better apply the results of identifying tea diseases,this study developed a mini-program that operates on WeChat.Users can quickly obtain accurate identification results and corresponding disease descriptions and prevention methods through simple operations.This intelligent tool for identifying tea diseases can serve as an auxiliary tool for farmers,consumers,and related scientific researchers and has certain practical value.
基金the support of the National Natural Science Foundation of China(Nos.12372190,12272077)the Fundamental Research Funds for the Central Universities,China(Nos.DUT20RC(5)009,DUT20GJ216).
文摘Good distribution of samples and weights can improve the computational accuracy and efficiency in the stochastic response analyses of aerospace problems with uncertain parameters.This work proposes a new Generalized L2 Discrepancy based on a General Point(GL2D-GP)for generating samples and their corresponding weights.The proposed GL2D-GP is an extension of the existing discrepancy by introducing the non-same weights and a smaller box to measure probability errors.Minimizing the GL2D-GP can yield a weight optimization formula that generates a set of optimal non-identical weights for a given sample set.Through minimizing the GL2D-GP assigned to the set of optimal non-same weights,a new sample and weight generation method is developed.In the proposed method,the samples can be easily generated in terms of the generalized Halton formula with a series of optimal permutation vectors which are found by the intelligent evolutionary algorithm.Once the sample set is obtained,the optimal weights can be generated in terms of the weight optimization formula.Five numerical examples are presented to verify the high accuracy,efficiency,and strong robustness of the proposed sample generation method based on GL2D-GP.
基金Sichuan Science and Technology Program under Grant No.2024NSFSC0932the National Natural Science Foundation of China under Grant No.52008047。
文摘Probabilistic assessment of seismic performance(SPPA)is a crucial aspect of evaluating the seismic behavior of structures.For complex bridges with inherent uncertainties,conducting precise and efficient seismic reliability analysis remains a significant challenge.To address this issue,the current study introduces a sample-unequal weight fractional moment assessment method,which is based on an improved correlation-reduced Latin hypercube sampling(ICLHS)technique.This method integrates the benefits of important sampling techniques with interpolator quadrature formulas to enhance the accuracy of estimating the extreme value distribution(EVD)for the seismic response of complex nonlinear structures subjected to non-stationary ground motions.Additionally,the core theoretical approaches employed in seismic reliability analysis(SRA)are elaborated,such as dimension reduction for simulating non-stationary random ground motions and a fractional-maximum entropy single-loop solution strategy.The effectiveness of this proposed method is validated through a three-story nonlinear shear frame structure.Furthermore,a comprehensive reliability analysis of a real-world long-span,single-pylon suspension bridge is conducted using the developed theoretical framework within the OpenSees platform,leading to key insights and conclusions.
文摘目的利用自适应合成抽样(adaptive synthetic sampling,ADASYN)与类别逆比例加权法处理类别不平衡数据,结合分类器构建模型对阿尔茨海默病(alzheimer′s disease,AD)患者疾病进程进行分类预测。方法数据源自阿尔茨海默病神经影像学计划(Alzheimer′s disease neuroimaging initiative,ADNI),经随机森林填补缺失值,弹性网络筛选特征子集后,利用ADASYN与类别逆比例加权法处理类别不平衡数据。分别结合随机森林(random forest,RF)、支持向量机(support vector machine,SVM)构建四种模型:ADASYN-RF、ADASYN-SVM、加权随机森林(weighted random forest,WRF)、加权支持向量机(weighted support vector machine,WSVM),与RF、SVM比较分类性能。模型评价指标为宏观平均精确率(macro-average of precision,macro-P)、宏观平均召回率(macro-average of recall,macro-R)、宏观平均F1值(macro-average of F1-score,macro-F1)、准确率(accuracy,ACC)、Kappa值和AUC(area under the ROC curve)。结果ADASYN-RF的分类性能最优(Kappa值为0.938,AUC为0.980),ADASYN-SVM次之。利用ADASYN-RF预测得到的重要分类特征分别为CDRSB、LDELTOTAL、MMSE,在临床上均可得到证实。结论ADASYN与类别逆比例加权法都能辅助提升分类器性能,但ADASYN算法更优。
