Defining the structure characteristics of amorphous materials is one of the fundamental problems that need to be solved urgently in complex materials because of their complex structure and long-range disorder.In this ...Defining the structure characteristics of amorphous materials is one of the fundamental problems that need to be solved urgently in complex materials because of their complex structure and long-range disorder.In this study,we develop an interpretable deep learning model capable of accurately classifying amorphous configurations and characterizing their structural properties.The results demonstrate that the multi-dimensional hybrid convolutional neural network can classify the two-dimensional(2D)liquids and amorphous solids of molecular dynamics simulation.The classification process does not make a priori assumptions on the amorphous particle environment,and the accuracy is 92.75%,which is better than other convolutional neural networks.Moreover,our model utilizes the gradient-weighted activation-like mapping method,which generates activation-like heat maps that can precisely identify important structures in the amorphous configuration maps.We obtain an order parameter from the heatmap and conduct finite scale analysis of this parameter.Our findings demonstrate that the order parameter effectively captures the amorphous phase transition process across various systems.These results hold significant scientific implications for the study of amorphous structural characteristics via deep learning.展开更多
Roof falls due to geological conditions are major hazards in the mining industry,causing work time loss,injuries,and fatalities.There are roof fall problems caused by high horizontal stress in several largeopening lim...Roof falls due to geological conditions are major hazards in the mining industry,causing work time loss,injuries,and fatalities.There are roof fall problems caused by high horizontal stress in several largeopening limestone mines in the eastern and midwestern United States.The typical hazard management approach for this type of roof fall hazards relies heavily on visual inspections and expert knowledge.In this context,we proposed a deep learning system for detection of the roof fall hazards caused by high horizontal stress.We used images depicting hazardous and non-hazardous roof conditions to develop a convolutional neural network(CNN)for autonomous detection of hazardous roof conditions.To compensate for limited input data,we utilized a transfer learning approach.In the transfer learning approach,an already-trained network is used as a starting point for classification in a similar domain.Results show that this approach works well for classifying roof conditions as hazardous or safe,achieving a statistical accuracy of 86.4%.This result is also compared with a random forest classifier,and the deep learning approach is more successful at classification of roof conditions.However,accuracy alone is not enough to ensure a reliable hazard management system.System constraints and reliability are improved when the features used by the network are understood.Therefore,we used a deep learning interpretation technique called integrated gradients to identify the important geological features in each image for prediction.The analysis of integrated gradients shows that the system uses the same roof features as the experts do on roof fall hazards detection.The system developed in this paper demonstrates the potential of deep learning in geotechnical hazard management to complement human experts,and likely to become an essential part of autonomous operations in cases where hazard identification heavily depends on expert knowledge.Moreover,deep learning-based systems reduce expert exposure to hazardous conditions.展开更多
Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the f...Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the focus of criticism,especially in the application of SARATR,which is closely associated with the national defense and security domain.To address these issues,a new interpretable recognition model Physics-Guided BagNet(PGBN)is proposed in this article.The model adopts an interpretable convolutional neural network framework and uses time–frequency analysis to extract physical scattering features in SAR images.Based on the physical scattering features,an unsupervised segmentation method is proposed to distinguish targets from the background in SAR images.On the basis of the segmentation result,a structure is designed,which constrains the model's spatial attention to focus more on the targets themselves rather than the background,thereby making the model's decision-making more in line with physical principles.In contrast to previous interpretable research methods,this model combines interpretable structure with physical interpretability,further reducing the model's risk of error recognition.Experiments on the MSTAR dataset verify that the PGBN model exhibits excellent interpretability and recognition performance,and comparative experiments with heatmaps indicate that the physical feature guidance module presented in this article can constrain the model to focus more on the target itself rather than the background.展开更多
With the development of information technology,radio communication technology has made rapid progress.Many radio signals that have appeared in space are difficult to classify without manually labeling.Unsupervised rad...With the development of information technology,radio communication technology has made rapid progress.Many radio signals that have appeared in space are difficult to classify without manually labeling.Unsupervised radio signal clustering methods have recently become an urgent need for this situation.Meanwhile,the high complexity of deep learning makes it difficult to understand the decision results of the clustering models,making it essential to conduct interpretable analysis.This paper proposed a combined loss function for unsupervised clustering based on autoencoder.The combined loss function includes reconstruction loss and deep clustering loss.Deep clustering loss is added based on reconstruction loss,which makes similar deep features converge more in feature space.In addition,a features visualization method for signal clustering was proposed to analyze the interpretability of autoencoder utilizing Saliency Map.Extensive experiments have been conducted on a modulated signal dataset,and the results indicate the superior performance of our proposed method over other clustering algorithms.In particular,for the simulated dataset containing six modulation modes,when the SNR is 20dB,the clustering accuracy of the proposed method is greater than 78%.The interpretability analysis of the clustering model was performed to visualize the significant features of different modulated signals and verified the high separability of the features extracted by clustering model.展开更多
Understanding the conformational characteristics of polymers is key to elucidating their physical properties.Cyclic polymers,defined by their closed-loop structures,inherently differ from linear polymers possessing di...Understanding the conformational characteristics of polymers is key to elucidating their physical properties.Cyclic polymers,defined by their closed-loop structures,inherently differ from linear polymers possessing distinct chain ends.Despite these structural differences,both types of polymers exhibit locally random-walk-like conformations,making it challenging to detect subtle spatial variations using conventional methods.In this study,we address this challenge by integrating molecular dynamics simulations with point cloud neural networks to analyze the spatial conformations of cyclic and linear polymers.By utilizing the Dynamic Graph CNN(DGCNN)model,we classify polymer conformations based on the 3D coordinates of monomers,capturing local and global topological differences without considering chain connectivity sequentiality.Our findings reveal that the optimal local structural feature unit size scales linearly with molecular weight,aligning with theoretical predictions.Additionally,interpretability techniques such as Grad-CAM and SHAP identify significant conformational differences:cyclic polymers tend to form prolate ellipsoid shapes with pronounced elongation along the major axis,while linear polymers show elongated ends with more spherical centers.These findings reveal subtle yet critical differences in local conformations between cyclic and linear polymers that were previously difficult to discern,providing deeper insights into polymer structure-property relationships and offering guidance for future polymer science advancements.展开更多
This paper reviews recent studies in understanding neural-network representations and learning neural networks with interpretable/disentangled middle-layer representations.Although deep neural networks have exhibited ...This paper reviews recent studies in understanding neural-network representations and learning neural networks with interpretable/disentangled middle-layer representations.Although deep neural networks have exhibited superior performance in various tasks,interpretability is always Achilles' heel of deep neural networks.At present,deep neural networks obtain high discrimination power at the cost of a low interpretability of their black-box representations.We believe that high model interpretability may help people break several bottlenecks of deep learning,e.g.,learning from a few annotations,learning via human–computer communications at the semantic level,and semantically debugging network representations.We focus on convolutional neural networks(CNNs),and revisit the visualization of CNN representations,methods of diagnosing representations of pre-trained CNNs,approaches for disentangling pre-trained CNN representations,learning of CNNs with disentangled representations,and middle-to-end learning based on model interpretability.Finally,we discuss prospective trends in explainable artificial intelligence.展开更多
To extract strong correlations between different energy loads and improve the interpretability and accuracy for load forecasting of a regional integrated energy system(RIES),an explainable framework for load forecasti...To extract strong correlations between different energy loads and improve the interpretability and accuracy for load forecasting of a regional integrated energy system(RIES),an explainable framework for load forecasting of an RIES is proposed.This includes the load forecasting model of RIES and its interpretation.A coupled feature extracting strat-egy is adopted to construct coupled features between loads as the input variables of the model.It is designed based on multi-task learning(MTL)with a long short-term memory(LSTM)model as the sharing layer.Based on SHapley Additive exPlanations(SHAP),this explainable framework combines global and local interpretations to improve the interpretability of load forecasting of the RIES.In addition,an input variable selection strategy based on the global SHAP value is proposed to select input feature variables of the model.A case study is given to verify the effectiveness of the proposed model,constructed coupled features,and input variable selection strategy.The results show that the explainable framework intuitively improves the interpretability of the prediction model.展开更多
This paper proposes a novel interpretable tourism demand forecasting framework that considers the impact of the COVID-19 pandemic by using multi-source heterogeneous data,namely,historical tourism volume,newly confirm...This paper proposes a novel interpretable tourism demand forecasting framework that considers the impact of the COVID-19 pandemic by using multi-source heterogeneous data,namely,historical tourism volume,newly confirmed cases in tourist origins and destinations,and search engine data.This paper introduces newly confirmed cases in tourist origins and tourist destinations to forecast tourism demand and proposes a new two-stage decomposition method called ensemble empirical mode decomposition-variational mode decomposition to deal with the tourist arrival sequence.To solve the problem of insufficient interpretability of existing tourism demand forecasting,this paper also proposes a novel interpretable tourism demand forecasting model called JADE-TFT,which utilizes an adaptive differential evolution algorithm with external archiving(JADE)to intelligently and efficiently optimize the hyperparameters of temporal fusion transformers(TFT).The validity of the proposed prediction framework is verified by actual cases based on Hainan and Macao tourism data sets.The interpretable experimental results show that newly confirmed cases in tourist origins and tourist destinations can better reflect tourists'concerns about travel in the post-pandemic era,and the two-stage decomposition method can effectively identify the inflection point of tourism prediction,thereby increasing the prediction accuracy of tourism demand.展开更多
基金National Natural Science Foundation of China(Grant No.11702289)the Key Core Technology and Generic Technology Research and Development Project of Shanxi Province,China(Grant No.2020XXX013)the National Key Research and Development Project of China。
文摘Defining the structure characteristics of amorphous materials is one of the fundamental problems that need to be solved urgently in complex materials because of their complex structure and long-range disorder.In this study,we develop an interpretable deep learning model capable of accurately classifying amorphous configurations and characterizing their structural properties.The results demonstrate that the multi-dimensional hybrid convolutional neural network can classify the two-dimensional(2D)liquids and amorphous solids of molecular dynamics simulation.The classification process does not make a priori assumptions on the amorphous particle environment,and the accuracy is 92.75%,which is better than other convolutional neural networks.Moreover,our model utilizes the gradient-weighted activation-like mapping method,which generates activation-like heat maps that can precisely identify important structures in the amorphous configuration maps.We obtain an order parameter from the heatmap and conduct finite scale analysis of this parameter.Our findings demonstrate that the order parameter effectively captures the amorphous phase transition process across various systems.These results hold significant scientific implications for the study of amorphous structural characteristics via deep learning.
基金partially supported by the National Institute for Occupational Safety and Health,contract number 0000HCCR-2019-36403。
文摘Roof falls due to geological conditions are major hazards in the mining industry,causing work time loss,injuries,and fatalities.There are roof fall problems caused by high horizontal stress in several largeopening limestone mines in the eastern and midwestern United States.The typical hazard management approach for this type of roof fall hazards relies heavily on visual inspections and expert knowledge.In this context,we proposed a deep learning system for detection of the roof fall hazards caused by high horizontal stress.We used images depicting hazardous and non-hazardous roof conditions to develop a convolutional neural network(CNN)for autonomous detection of hazardous roof conditions.To compensate for limited input data,we utilized a transfer learning approach.In the transfer learning approach,an already-trained network is used as a starting point for classification in a similar domain.Results show that this approach works well for classifying roof conditions as hazardous or safe,achieving a statistical accuracy of 86.4%.This result is also compared with a random forest classifier,and the deep learning approach is more successful at classification of roof conditions.However,accuracy alone is not enough to ensure a reliable hazard management system.System constraints and reliability are improved when the features used by the network are understood.Therefore,we used a deep learning interpretation technique called integrated gradients to identify the important geological features in each image for prediction.The analysis of integrated gradients shows that the system uses the same roof features as the experts do on roof fall hazards detection.The system developed in this paper demonstrates the potential of deep learning in geotechnical hazard management to complement human experts,and likely to become an essential part of autonomous operations in cases where hazard identification heavily depends on expert knowledge.Moreover,deep learning-based systems reduce expert exposure to hazardous conditions.
基金co-supported by the National Natural Science Foundation of China(No.62001507)the Youth Talent Lifting Project of the China Association for Science and Technology(No.2021-JCJQ-QT-018)+1 种基金the Program of the Youth Innovation Team of Shaanxi Universitiesthe Natural Science Basic Research Plan in Shaanxi Province of China(No.2023-JC-YB-491)。
文摘Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the focus of criticism,especially in the application of SARATR,which is closely associated with the national defense and security domain.To address these issues,a new interpretable recognition model Physics-Guided BagNet(PGBN)is proposed in this article.The model adopts an interpretable convolutional neural network framework and uses time–frequency analysis to extract physical scattering features in SAR images.Based on the physical scattering features,an unsupervised segmentation method is proposed to distinguish targets from the background in SAR images.On the basis of the segmentation result,a structure is designed,which constrains the model's spatial attention to focus more on the targets themselves rather than the background,thereby making the model's decision-making more in line with physical principles.In contrast to previous interpretable research methods,this model combines interpretable structure with physical interpretability,further reducing the model's risk of error recognition.Experiments on the MSTAR dataset verify that the PGBN model exhibits excellent interpretability and recognition performance,and comparative experiments with heatmaps indicate that the physical feature guidance module presented in this article can constrain the model to focus more on the target itself rather than the background.
基金supported in part by the National Natural Science Foundation of China(No.62276206)the Key Research and Development Program of Shaanxi under Grant S2022-YF-YBGY-0921+2 种基金the State Key Program of National Natural Science of China(No.62231027)supported by the Science and Technology on Communication Information Security Control Laboratory。
文摘With the development of information technology,radio communication technology has made rapid progress.Many radio signals that have appeared in space are difficult to classify without manually labeling.Unsupervised radio signal clustering methods have recently become an urgent need for this situation.Meanwhile,the high complexity of deep learning makes it difficult to understand the decision results of the clustering models,making it essential to conduct interpretable analysis.This paper proposed a combined loss function for unsupervised clustering based on autoencoder.The combined loss function includes reconstruction loss and deep clustering loss.Deep clustering loss is added based on reconstruction loss,which makes similar deep features converge more in feature space.In addition,a features visualization method for signal clustering was proposed to analyze the interpretability of autoencoder utilizing Saliency Map.Extensive experiments have been conducted on a modulated signal dataset,and the results indicate the superior performance of our proposed method over other clustering algorithms.In particular,for the simulated dataset containing six modulation modes,when the SNR is 20dB,the clustering accuracy of the proposed method is greater than 78%.The interpretability analysis of the clustering model was performed to visualize the significant features of different modulated signals and verified the high separability of the features extracted by clustering model.
基金the National Key R&D Program of China(No.2022YFB3707303)National Natural Science Foundation of China(No.52293471)。
文摘Understanding the conformational characteristics of polymers is key to elucidating their physical properties.Cyclic polymers,defined by their closed-loop structures,inherently differ from linear polymers possessing distinct chain ends.Despite these structural differences,both types of polymers exhibit locally random-walk-like conformations,making it challenging to detect subtle spatial variations using conventional methods.In this study,we address this challenge by integrating molecular dynamics simulations with point cloud neural networks to analyze the spatial conformations of cyclic and linear polymers.By utilizing the Dynamic Graph CNN(DGCNN)model,we classify polymer conformations based on the 3D coordinates of monomers,capturing local and global topological differences without considering chain connectivity sequentiality.Our findings reveal that the optimal local structural feature unit size scales linearly with molecular weight,aligning with theoretical predictions.Additionally,interpretability techniques such as Grad-CAM and SHAP identify significant conformational differences:cyclic polymers tend to form prolate ellipsoid shapes with pronounced elongation along the major axis,while linear polymers show elongated ends with more spherical centers.These findings reveal subtle yet critical differences in local conformations between cyclic and linear polymers that were previously difficult to discern,providing deeper insights into polymer structure-property relationships and offering guidance for future polymer science advancements.
基金supported by the ONR MURI pro ject(No.N00014-16-1-2007)the DARPA XAI Award(No.N66001-17-2-4029)NSF IIS(No.1423305)
文摘This paper reviews recent studies in understanding neural-network representations and learning neural networks with interpretable/disentangled middle-layer representations.Although deep neural networks have exhibited superior performance in various tasks,interpretability is always Achilles' heel of deep neural networks.At present,deep neural networks obtain high discrimination power at the cost of a low interpretability of their black-box representations.We believe that high model interpretability may help people break several bottlenecks of deep learning,e.g.,learning from a few annotations,learning via human–computer communications at the semantic level,and semantically debugging network representations.We focus on convolutional neural networks(CNNs),and revisit the visualization of CNN representations,methods of diagnosing representations of pre-trained CNNs,approaches for disentangling pre-trained CNN representations,learning of CNNs with disentangled representations,and middle-to-end learning based on model interpretability.Finally,we discuss prospective trends in explainable artificial intelligence.
基金supported in part by the National Key Research Program of China (2016YFB0900100)Key Project of Shanghai Science and Technology Committee (18DZ1100303).
文摘To extract strong correlations between different energy loads and improve the interpretability and accuracy for load forecasting of a regional integrated energy system(RIES),an explainable framework for load forecasting of an RIES is proposed.This includes the load forecasting model of RIES and its interpretation.A coupled feature extracting strat-egy is adopted to construct coupled features between loads as the input variables of the model.It is designed based on multi-task learning(MTL)with a long short-term memory(LSTM)model as the sharing layer.Based on SHapley Additive exPlanations(SHAP),this explainable framework combines global and local interpretations to improve the interpretability of load forecasting of the RIES.In addition,an input variable selection strategy based on the global SHAP value is proposed to select input feature variables of the model.A case study is given to verify the effectiveness of the proposed model,constructed coupled features,and input variable selection strategy.The results show that the explainable framework intuitively improves the interpretability of the prediction model.
基金partially supported by the Humanities and Social Sciences Foundation of the Chinese Ministry of Education of China under Grant No.22YJA630003。
文摘This paper proposes a novel interpretable tourism demand forecasting framework that considers the impact of the COVID-19 pandemic by using multi-source heterogeneous data,namely,historical tourism volume,newly confirmed cases in tourist origins and destinations,and search engine data.This paper introduces newly confirmed cases in tourist origins and tourist destinations to forecast tourism demand and proposes a new two-stage decomposition method called ensemble empirical mode decomposition-variational mode decomposition to deal with the tourist arrival sequence.To solve the problem of insufficient interpretability of existing tourism demand forecasting,this paper also proposes a novel interpretable tourism demand forecasting model called JADE-TFT,which utilizes an adaptive differential evolution algorithm with external archiving(JADE)to intelligently and efficiently optimize the hyperparameters of temporal fusion transformers(TFT).The validity of the proposed prediction framework is verified by actual cases based on Hainan and Macao tourism data sets.The interpretable experimental results show that newly confirmed cases in tourist origins and tourist destinations can better reflect tourists'concerns about travel in the post-pandemic era,and the two-stage decomposition method can effectively identify the inflection point of tourism prediction,thereby increasing the prediction accuracy of tourism demand.
