Building accurate and generalizable machine-learning models requires large training datasets.In aerodynamics,quantities of interest are typically governed by complex,non-linear mechanisms in which neural networks are ...Building accurate and generalizable machine-learning models requires large training datasets.In aerodynamics,quantities of interest are typically governed by complex,non-linear mechanisms in which neural networks are well-suited to address.However,the acquisition of large,high-fidelity datasets from either simulations or experiments can be expensive.In this work,a transfer-learning framework is explored to reduce the reliance on these expensive datasets by exploiting the cost-effectiveness of lowfidelity analyses in constructing extensive datasets,such as the inviscid panel method.By frst developing robust base networks from inviscid distributions,target networks can"learn"by simply transferring relevant embedded features to facilitate the modelling of high-fidelity distributions,instead of solely relying on its access to high-fidelity samples.Assessment of the framework reveals performance gains over conventional training schemes in(1)fidelity enhancement from inviscid to high-fidelity pressure distributions;(2)generalizing prior knowledge to learn adjacent skin friction properties even without a low-fidelity equivalent;(3)extrapolation to yet-to-be seen operating conditions.Under conditions of limited high-fidelity samples,test MSE evaluations can be improved by magnitudes of up to 10^(2),10^(1),and 10^(2) for the three respective tasks.As such,these findings motivate further investigations to support data-scarce surrogate modelling in more empirical settings.展开更多
(1) Machine learning methods to assist energy system optimization,by A.T.D.Perera,P.U.Wickramasinghe,Vahid M.Nik,Jean-Louis Scartezzini,Pages 191-205 Abstract: This study evaluates the potential of supervised and tran...(1) Machine learning methods to assist energy system optimization,by A.T.D.Perera,P.U.Wickramasinghe,Vahid M.Nik,Jean-Louis Scartezzini,Pages 191-205 Abstract: This study evaluates the potential of supervised and transfer learning techniques to assist energy system optimization.A surrogate model is developed with the support of a supervised learning technique (by using artificial neural network) in order to bypass computationally intensive Actual Engineering Model (AEM).Eight different neural network architectures are considered in the process of developing the surrogate model.Subsequently,a hybrid optimization algorithm (HOA) is developed combining Surrogate and AEM in order to speed up the optimization process while maintaining the accuracy.Pareto optimization is conducted considering Net Present Value and Grid Integration level as the objective functions.Transfer learning is used to adapt the surrogate model (trained using supervised learning technique) for different scenarios where solar energy potential,wind speed and energy demand are notably different.Results reveal that the surrogate model can reach to Pareto solutions with a higher accuracy when grid interactions are above 10 %(with reasonable differences in the decision space variables).HOA can reach to Pareto solutions (similar to the solutions obtained using AEM) around 17 times faster than AEM.The Surrogate Models developed using Transfer Learning (SMTL) shows a similar capability.SMTL combined with the optimization algorithm can predict Pareto fronts efficiently even when there are significant changes in the initial conditions.Therefore,STML can be used along with the HOA,which reduces the computational time required for energy system optimization by 84 %.Such a significant reduction in computational time enables the approach to be used for energy system optimization at regional or national scale.展开更多
Background:This study aims to predict the extraction difficulty of mandibular third molars based on panoramic images using transfer learning while employing super-resolution(SR)technology to enhance the feasibility an...Background:This study aims to predict the extraction difficulty of mandibular third molars based on panoramic images using transfer learning while employing super-resolution(SR)technology to enhance the feasibility and validity of the prediction.Methods:We reviewed a total of 608 preoperative mandibular third molar panoramic radiographs from two medical facilities:the First Affiliated Hospital of Zhengzhou University(n=509;456 in the training set and 53 in the test set)and the Henan Provincial Dental Hospital(n=99 in the validation set).We conducted a deep-transfer learning network on high-resolution(HR)panoramic radiographs to improve the longitudinal resolution of the images and obtained the SR images.Subsequently,we constructed models named Model-HR and Model-SR using high-dimensional quantitative features extracted through the Least Absolute Shrinkage and Selection Operator method.The models’performances were evaluated using the receiver operating characteristic curve(ROC).To assess the reliability of the model,we compared the results from the test set with those of three dentists.Results:Model-SR outperformed Model-HR(area under the curve(AUC):0.779,sensitivity:85.5%,specificity:60.9%,and accuracy:79.8%vs.AUC:0.753,sensitivity:73.7%,specificity:73.9%,and accuracy:73.7%)in predicting the difficulty of extracting mandibular third molars.Both Model-HR(AUC=0.821,95%CI 0.687–0.956)and Model-SR(AUC=0.963,95%CI 0.921–0.999)demonstrated superior performance compared to expert dentists(highest AUC=0.799,95%CI 0.671–0.927).Conclusions:Model-SR yielded superior predictive performance in determining the difficulty of extracting mandibular third molars when compared with Model-HR and expert dentists’visual assessments.展开更多
基金conducted during Benjamin YJ's graduate studies at the Department of Mechanical Engineering,National University of Singapore underthe support of the NUS-DSO Graduate Programme.
文摘Building accurate and generalizable machine-learning models requires large training datasets.In aerodynamics,quantities of interest are typically governed by complex,non-linear mechanisms in which neural networks are well-suited to address.However,the acquisition of large,high-fidelity datasets from either simulations or experiments can be expensive.In this work,a transfer-learning framework is explored to reduce the reliance on these expensive datasets by exploiting the cost-effectiveness of lowfidelity analyses in constructing extensive datasets,such as the inviscid panel method.By frst developing robust base networks from inviscid distributions,target networks can"learn"by simply transferring relevant embedded features to facilitate the modelling of high-fidelity distributions,instead of solely relying on its access to high-fidelity samples.Assessment of the framework reveals performance gains over conventional training schemes in(1)fidelity enhancement from inviscid to high-fidelity pressure distributions;(2)generalizing prior knowledge to learn adjacent skin friction properties even without a low-fidelity equivalent;(3)extrapolation to yet-to-be seen operating conditions.Under conditions of limited high-fidelity samples,test MSE evaluations can be improved by magnitudes of up to 10^(2),10^(1),and 10^(2) for the three respective tasks.As such,these findings motivate further investigations to support data-scarce surrogate modelling in more empirical settings.
文摘(1) Machine learning methods to assist energy system optimization,by A.T.D.Perera,P.U.Wickramasinghe,Vahid M.Nik,Jean-Louis Scartezzini,Pages 191-205 Abstract: This study evaluates the potential of supervised and transfer learning techniques to assist energy system optimization.A surrogate model is developed with the support of a supervised learning technique (by using artificial neural network) in order to bypass computationally intensive Actual Engineering Model (AEM).Eight different neural network architectures are considered in the process of developing the surrogate model.Subsequently,a hybrid optimization algorithm (HOA) is developed combining Surrogate and AEM in order to speed up the optimization process while maintaining the accuracy.Pareto optimization is conducted considering Net Present Value and Grid Integration level as the objective functions.Transfer learning is used to adapt the surrogate model (trained using supervised learning technique) for different scenarios where solar energy potential,wind speed and energy demand are notably different.Results reveal that the surrogate model can reach to Pareto solutions with a higher accuracy when grid interactions are above 10 %(with reasonable differences in the decision space variables).HOA can reach to Pareto solutions (similar to the solutions obtained using AEM) around 17 times faster than AEM.The Surrogate Models developed using Transfer Learning (SMTL) shows a similar capability.SMTL combined with the optimization algorithm can predict Pareto fronts efficiently even when there are significant changes in the initial conditions.Therefore,STML can be used along with the HOA,which reduces the computational time required for energy system optimization by 84 %.Such a significant reduction in computational time enables the approach to be used for energy system optimization at regional or national scale.
基金supported by the National Natural Science Foundation of China(U1904145)the Joint Funds for the Innovation of Science and Technology of Fujian province(2019Y9128).
文摘Background:This study aims to predict the extraction difficulty of mandibular third molars based on panoramic images using transfer learning while employing super-resolution(SR)technology to enhance the feasibility and validity of the prediction.Methods:We reviewed a total of 608 preoperative mandibular third molar panoramic radiographs from two medical facilities:the First Affiliated Hospital of Zhengzhou University(n=509;456 in the training set and 53 in the test set)and the Henan Provincial Dental Hospital(n=99 in the validation set).We conducted a deep-transfer learning network on high-resolution(HR)panoramic radiographs to improve the longitudinal resolution of the images and obtained the SR images.Subsequently,we constructed models named Model-HR and Model-SR using high-dimensional quantitative features extracted through the Least Absolute Shrinkage and Selection Operator method.The models’performances were evaluated using the receiver operating characteristic curve(ROC).To assess the reliability of the model,we compared the results from the test set with those of three dentists.Results:Model-SR outperformed Model-HR(area under the curve(AUC):0.779,sensitivity:85.5%,specificity:60.9%,and accuracy:79.8%vs.AUC:0.753,sensitivity:73.7%,specificity:73.9%,and accuracy:73.7%)in predicting the difficulty of extracting mandibular third molars.Both Model-HR(AUC=0.821,95%CI 0.687–0.956)and Model-SR(AUC=0.963,95%CI 0.921–0.999)demonstrated superior performance compared to expert dentists(highest AUC=0.799,95%CI 0.671–0.927).Conclusions:Model-SR yielded superior predictive performance in determining the difficulty of extracting mandibular third molars when compared with Model-HR and expert dentists’visual assessments.
