Metal–organic frameworks(MOFs)hold great potential for gas separation and storage,and graph neural networks have proven to be a powerful tool for exploring material structure–property relationships and discovering n...Metal–organic frameworks(MOFs)hold great potential for gas separation and storage,and graph neural networks have proven to be a powerful tool for exploring material structure–property relationships and discovering new materials.Unlike traditional molecular graphs,crystal graphs require consideration of periodic invariance and modes.In addition,MOF structures such as covalent bonds,functional groups,and global structures impact adsorption performance in different ways.However,redundant atomic interactions can disrupt training accuracy,potentially leading to overfitting.In this paper,we propose a multi-scale crystal graph for describing periodic crystal structures,modeling interatomic interactions at different scales while preserving periodicity invariance.We also propose a multi-head attention crystal graph network in multi-scale graphs(MHACGN-MS),which learns structural characteristics by focusing on interatomic interactions at different scales,thereby reducing interference from redundant interactions.Using MOF adsorption for gases as an example,we demonstrate that MHACGN-MS outperforms traditional graph neural networks in predicting multi-component gas adsorption.We also visualize attention scores to validate effective learning and demonstrate the model’s interpretability.展开更多
Subcellular location is one of the key biological characteristics of proteins. Position-specific profiles （PSP） have been introduced as important characteristics of proteins in this article. In this study, to obtain...Subcellular location is one of the key biological characteristics of proteins. Position-specific profiles （PSP） have been introduced as important characteristics of proteins in this article. In this study, to obtain position-specific profiles, the Position Specific lterative-Basic Local Alignment Search Tool （PSI-BLAST） has been used to search for protein sequences in a database. Position-specific scoring matrices are extracted from the profiles as one class of characteristics. Four-part amino acid compositions and lst-7th order dipeptide compositions have also been calculated as the other two classes of characteristics. Therefore, twelve characteristic vectors are extracted from each of the protein sequences. Next, the characteristic vectors are weighed by a simple weighing function and inputted into a BP neural network predictor named PSP-Weighted Neural Network （PSP-WNN）. The Levenberg-Marquardt algorithm is employed to adjust the weight matrices and thresholds during the network training instead of the error back propagation algorithm. With a jackknife test on the RH2427 dataset, PSP-WNN has achieved a higher overall prediction accuracy of 88.4% rather than the prediction results by the general BP neural network, Markov model, and fuzzy k-nearest neighbors algorithm on this dataset. In addition, the prediction performance of PSP-WNN has been evaluated with a five-fold cross validation test on the PK7579 dataset and the prediction results have been consistently better than those of the previous method on the basis of several support vector machines, using compositions of both amino acids and amino acid pairs. These results indicate that PSP-WNN is a powerful tool for subcellular localization prediction. At the end of the article, influences on prediction accuracy using different weighting proportions among three characteristic vector categories have been discussed. An appropriate proportion is considered by increasing the prediction accuracy.展开更多
Fluid-attenuated inversion recovery （FLAIR） vascular hyperintensity （FVH） is used to assess leptomeningeal collateral circulation, but clinical outcomes of patients with FVH can be very different. The aim of the p...Fluid-attenuated inversion recovery （FLAIR） vascular hyperintensity （FVH） is used to assess leptomeningeal collateral circulation, but clinical outcomes of patients with FVH can be very different. The aim of the present study was to assess a FVH score and explore its relationship with clinical outcomes. Patients with acute ischemic stroke due to middle cerebral artery M1 occlusion underwent magnetic resonance imaging and were followed up at 10 days （National Institutes of Health Stroke Scale） and 90 days （modified Rankin Scale） to determine short-term clinical outcomes. Effective collateral circulation indirectly improved recovery of neurological function and short-term clinical outcome by extending the size of the pial penumbra and reducing infarct lesions. FVH score showed no correlation with 90-day functional clinical outcome and was not sufficient as an independent predictor of short-term clinical outcome.展开更多
Exploring the relationship between different structure of the spinal cord and functional assessment after spinal cord injury is important. Quantitative diffusion tensor imaging can provide information about the micros...Exploring the relationship between different structure of the spinal cord and functional assessment after spinal cord injury is important. Quantitative diffusion tensor imaging can provide information about the microstructure of nerve tissue and can quantify the pathological damage of spinal cord white matter and gray matter. In this study, a custom-designed spinal cord contusion-impactor was used to damage the T_（10） spinal cord of beagles. Diffusion tensor imaging was used to observe changes in the whole spinal cord, white matter, and gray matter, and the Texas Spinal Cord Injury Score was used to assess changes in neurological function at 3 hours, 24 hours, 6 weeks, and 12 weeks after injury. With time, fractional anisotropy values after spinal cord injury showed a downward trend, and the apparent diffusion coefficient, mean diffusivity, and radial diffusivity first decreased and then increased. The apparent diffusion-coefficient value was highly associated with the Texas Spinal Cord Injury Score for the whole spinal cord（R = 0.919, P = 0.027）, white matter（R = 0.932, P = 0.021）, and gray matter（R = 0.882, P = 0.048）. Additionally, the other parameters had almost no correlation with the score（P 〉 0.05）. In conclusion, the highest and most significant correlation between diffusion parameters and neurological function was the apparent diffusion-coefficient value for white matter, indicating that it could be used to predict the recovery of neurological function accurately after spinal cord injury.展开更多
OBJECTIVES:To observe recovery in movement function in rats with middle cerebral artery occlusion(MCAO) after acupuncture treatment.METHODS:According to the randomized and controlled principle 1384 rats were divided i...OBJECTIVES:To observe recovery in movement function in rats with middle cerebral artery occlusion(MCAO) after acupuncture treatment.METHODS:According to the randomized and controlled principle 1384 rats were divided into the ba-sic control group(including the normal,sham,model control,model without intervention,Nimodipine,and para-Renzhong groups) and the acupuncture group(including the Neiguan(PC 6),Weizhong(BL 40),Sanyinjiao(SP 6),Chize(LU 5),Renzhong(GV 6) and non-acupoint groups).MCAO was modeled by Zea-longa's thread ligation and rats with scores of 1-3,as assessed by Zausinger's six-point method,were used in this study.Moreover,in the acupuncture group each acupoint was set with 12 different parameters by the orthogonal intersection method,resulting in 78 groups with 18 rats per group.The rats were treated by acupuncture once every 12 h for a total of six sessions and neurobehavioral scores were measured after each session.The neurobehavioral scores were compared by one-way ANOVA using the statistical software SPSS 17.0.RESULTS:After acupuncture therapy the mean neurobehavioral scores in MCAO rats increased gradually at each time point with a significant difference among the six scores,but with no significant differences between the fourth(48 h) and the fifth score(60 h),and between the fifth(60 h) and the sixth(72 h) score(P > 0.05).CONCLUSIONS:MCAO rats gradually recovered movement function over multiple acupuncture sessions.After the fouth acupuncture session(48 h),the neurobehavioral scores of rats with cerebral infarction remained stable.Acupuncture treatment had a reliable curative effect on movement function in cerebral infarction rats.展开更多
This paper describes the application of principal component analysis (PCA) and artificial neural network (ANN) to predict the air pollutant index (API) within the seven selected Malaysian air monitoring stations in th...This paper describes the application of principal component analysis (PCA) and artificial neural network (ANN) to predict the air pollutant index (API) within the seven selected Malaysian air monitoring stations in the southern region of Peninsular Malaysia based on seven years database (2005-2011). Feed-forward ANN was used as a prediction method. The feed-forward ANN analysis demonstrated that the rotated principal component scores (RPCs) were the best input parameters to predict API. From the 4 RPCs, only 10 (CO, O3, PM10, NO2, CH4, NmHC, THC, wind direction, humidity and ambient temp) out of 12 prediction variables were the most significant parameters to predict API. The results proved that the ANN method can be applied successfully as tools for decision making and problem solving for better atmospheric management.展开更多
A multimodal biometric system is applied to recognize individuals for authentication using neural networks. In this paper multimodal biometric algorithm is designed by integrating iris, finger vein, palm print and fac...A multimodal biometric system is applied to recognize individuals for authentication using neural networks. In this paper multimodal biometric algorithm is designed by integrating iris, finger vein, palm print and face biometric traits. Normalized score level fusion approach is applied and optimized, encoded for matching decision. It is a multilevel wavelet, phase based fusion algorithm. This robust multimodal biometric algorithm increases the security level, accuracy, reduces memory size and equal error rate and eliminates unimodal biometric algorithm vulnerabilities.展开更多
With the rapid development of the movie industry, it is vital to evaluate and predict a movie’s quality. In this paper, a movie score prediction model is proposed based on the movie plots. Movie data was processed wi...With the rapid development of the movie industry, it is vital to evaluate and predict a movie’s quality. In this paper, a movie score prediction model is proposed based on the movie plots. Movie data was processed with the word2vec method, and the linear regression model and back propagation neural network algorithm were employed to establish the movie score prediction model. The high-quality classic movie plots of high-scoring movies summed up by big data contributed to a high synthesis of the wonderful content of the film. Experimental results show that it is effective in terms of movie evaluation and prediction, and helpful in understanding people’s preferences for movie plots.展开更多
Bank credit is one of the effective ways to meet the financing needs of small and micro enterprises and support the development of inclusive finance. In the credit audit of small and micro enterprises, quantitative ri...Bank credit is one of the effective ways to meet the financing needs of small and micro enterprises and support the development of inclusive finance. In the credit audit of small and micro enterprises, quantitative risk assessment based on statistical methods plays an increasingly important role. With the development of computer technology and artificial intelligence technology, enterprise general credit evaluation technology is gradually exploring from traditional method to deep learning algorithm. This paper introduces the method of modeling the credit score of small and micro enterprises based on the neural network algorithm, and according to the empirical results, verifies the feasibility of the application of the deep learning algorithm in the credit score of small and micro enterprises.展开更多
This study constructs a proposed model to investigate the link between environmental,social,and governance(ESG)disclosures and ESG scores for publicly traded companies in the Borsa Istanbul Sustainability(XUSRD)index....This study constructs a proposed model to investigate the link between environmental,social,and governance(ESG)disclosures and ESG scores for publicly traded companies in the Borsa Istanbul Sustainability(XUSRD)index.In this context,this study considers 66 companies,examining recently structured ESG disclosures for 2022 that were published for the first time as novel data and applying a multilayer perceptron(MLP)artificial neural network algorithm.The relevant results are fourfold.(1)The MLP algorithm has explanatory power(i.e.,R^(2))of 79% in estimating companies’ESG scores.(2)Common,environment,social,and governance pillars have respective weights of 21.04%,44.87%,30.34%,and 3.74% in total ESG scores.(3)The absolute and relative significance of each ESG reporting principle for companies’ESG scores varies.(4)According to absolute and relative significance,the most effective ESG principle is the common principle,followed by social and environmental principles,whereas governance principles have less significance.Overall,the results demonstrate that applying a linear approach to complete deficient ESG disclosures is inefficient for increasing companies’ESG scores;instead,companies should focus on the ESG principles that have the highest relative significance.The findings of this study contribute to the literature by defining the most significant ESG principles for stimulating the ESG scores of companies in the XUSRD index.展开更多
Peer-to-peer (P2P) lending offers an alternative way to access credit. Unlike established lending institutions with proven credit risk management practices, P2P platforms rely on numerous independent variables to eval...Peer-to-peer (P2P) lending offers an alternative way to access credit. Unlike established lending institutions with proven credit risk management practices, P2P platforms rely on numerous independent variables to evaluate loan applicants’ creditworthiness. This study aims to estimate default probabilities using a mixture-of-experts neural network in P2P lending. The approach involves coupling unsupervised clustering to capture essential data properties with a classification algorithm based on the mixture-of-experts structure. This classic design enhances model capacity without significant computational overhead. The model was tested using P2P data from Lending Club, comparing it to other methods like Logistic Regression, AdaBoost, Gradient Boosting, Decision Tree, Support Vector Machine, and Random Forest. The hybrid model demonstrated superior performance, with a Mean Squared Error reduction of at least 25%.展开更多
基金supported by the National Natural Science Foundation of China(NSFC)(61821005).
文摘Metal–organic frameworks(MOFs)hold great potential for gas separation and storage,and graph neural networks have proven to be a powerful tool for exploring material structure–property relationships and discovering new materials.Unlike traditional molecular graphs,crystal graphs require consideration of periodic invariance and modes.In addition,MOF structures such as covalent bonds,functional groups,and global structures impact adsorption performance in different ways.However,redundant atomic interactions can disrupt training accuracy,potentially leading to overfitting.In this paper,we propose a multi-scale crystal graph for describing periodic crystal structures,modeling interatomic interactions at different scales while preserving periodicity invariance.We also propose a multi-head attention crystal graph network in multi-scale graphs(MHACGN-MS),which learns structural characteristics by focusing on interatomic interactions at different scales,thereby reducing interference from redundant interactions.Using MOF adsorption for gases as an example,we demonstrate that MHACGN-MS outperforms traditional graph neural networks in predicting multi-component gas adsorption.We also visualize attention scores to validate effective learning and demonstrate the model’s interpretability.
基金the National Natural Science Foundation of China (No. 60471003).
文摘Subcellular location is one of the key biological characteristics of proteins. Position-specific profiles （PSP） have been introduced as important characteristics of proteins in this article. In this study, to obtain position-specific profiles, the Position Specific lterative-Basic Local Alignment Search Tool （PSI-BLAST） has been used to search for protein sequences in a database. Position-specific scoring matrices are extracted from the profiles as one class of characteristics. Four-part amino acid compositions and lst-7th order dipeptide compositions have also been calculated as the other two classes of characteristics. Therefore, twelve characteristic vectors are extracted from each of the protein sequences. Next, the characteristic vectors are weighed by a simple weighing function and inputted into a BP neural network predictor named PSP-Weighted Neural Network （PSP-WNN）. The Levenberg-Marquardt algorithm is employed to adjust the weight matrices and thresholds during the network training instead of the error back propagation algorithm. With a jackknife test on the RH2427 dataset, PSP-WNN has achieved a higher overall prediction accuracy of 88.4% rather than the prediction results by the general BP neural network, Markov model, and fuzzy k-nearest neighbors algorithm on this dataset. In addition, the prediction performance of PSP-WNN has been evaluated with a five-fold cross validation test on the PK7579 dataset and the prediction results have been consistently better than those of the previous method on the basis of several support vector machines, using compositions of both amino acids and amino acid pairs. These results indicate that PSP-WNN is a powerful tool for subcellular localization prediction. At the end of the article, influences on prediction accuracy using different weighting proportions among three characteristic vector categories have been discussed. An appropriate proportion is considered by increasing the prediction accuracy.
基金supported by the National Natural Science Foundation of China,No.81371521
文摘Fluid-attenuated inversion recovery （FLAIR） vascular hyperintensity （FVH） is used to assess leptomeningeal collateral circulation, but clinical outcomes of patients with FVH can be very different. The aim of the present study was to assess a FVH score and explore its relationship with clinical outcomes. Patients with acute ischemic stroke due to middle cerebral artery M1 occlusion underwent magnetic resonance imaging and were followed up at 10 days （National Institutes of Health Stroke Scale） and 90 days （modified Rankin Scale） to determine short-term clinical outcomes. Effective collateral circulation indirectly improved recovery of neurological function and short-term clinical outcome by extending the size of the pial penumbra and reducing infarct lesions. FVH score showed no correlation with 90-day functional clinical outcome and was not sufficient as an independent predictor of short-term clinical outcome.
基金supported by the National Natural Science Foundation of China,No.81272164the Special Fund for Basic Scientific Research of Central Public Research Institutes in China,No.2015CZ-6,2016CZ-4+2 种基金the Beijing Institute for Brain Disorders in China,No.201601,0000-100031the Supporting Program of the “Twelve Five-year Plan” for Science&Technology Research of China,No.2012BAI34B02a grant from the Ministry of Science and Technology of China,No.2015CB351701
文摘Exploring the relationship between different structure of the spinal cord and functional assessment after spinal cord injury is important. Quantitative diffusion tensor imaging can provide information about the microstructure of nerve tissue and can quantify the pathological damage of spinal cord white matter and gray matter. In this study, a custom-designed spinal cord contusion-impactor was used to damage the T_（10） spinal cord of beagles. Diffusion tensor imaging was used to observe changes in the whole spinal cord, white matter, and gray matter, and the Texas Spinal Cord Injury Score was used to assess changes in neurological function at 3 hours, 24 hours, 6 weeks, and 12 weeks after injury. With time, fractional anisotropy values after spinal cord injury showed a downward trend, and the apparent diffusion coefficient, mean diffusivity, and radial diffusivity first decreased and then increased. The apparent diffusion-coefficient value was highly associated with the Texas Spinal Cord Injury Score for the whole spinal cord（R = 0.919, P = 0.027）, white matter（R = 0.932, P = 0.021）, and gray matter（R = 0.882, P = 0.048）. Additionally, the other parameters had almost no correlation with the score（P 〉 0.05）. In conclusion, the highest and most significant correlation between diffusion parameters and neurological function was the apparent diffusion-coefficient value for white matter, indicating that it could be used to predict the recovery of neurological function accurately after spinal cord injury.
基金Supported by Development Plan (973 Plan) of National Critical and Basic Research:2006 CB 504504,2010 CB 530500
文摘OBJECTIVES:To observe recovery in movement function in rats with middle cerebral artery occlusion(MCAO) after acupuncture treatment.METHODS:According to the randomized and controlled principle 1384 rats were divided into the ba-sic control group(including the normal,sham,model control,model without intervention,Nimodipine,and para-Renzhong groups) and the acupuncture group(including the Neiguan(PC 6),Weizhong(BL 40),Sanyinjiao(SP 6),Chize(LU 5),Renzhong(GV 6) and non-acupoint groups).MCAO was modeled by Zea-longa's thread ligation and rats with scores of 1-3,as assessed by Zausinger's six-point method,were used in this study.Moreover,in the acupuncture group each acupoint was set with 12 different parameters by the orthogonal intersection method,resulting in 78 groups with 18 rats per group.The rats were treated by acupuncture once every 12 h for a total of six sessions and neurobehavioral scores were measured after each session.The neurobehavioral scores were compared by one-way ANOVA using the statistical software SPSS 17.0.RESULTS:After acupuncture therapy the mean neurobehavioral scores in MCAO rats increased gradually at each time point with a significant difference among the six scores,but with no significant differences between the fourth(48 h) and the fifth score(60 h),and between the fifth(60 h) and the sixth(72 h) score(P > 0.05).CONCLUSIONS:MCAO rats gradually recovered movement function over multiple acupuncture sessions.After the fouth acupuncture session(48 h),the neurobehavioral scores of rats with cerebral infarction remained stable.Acupuncture treatment had a reliable curative effect on movement function in cerebral infarction rats.
文摘This paper describes the application of principal component analysis (PCA) and artificial neural network (ANN) to predict the air pollutant index (API) within the seven selected Malaysian air monitoring stations in the southern region of Peninsular Malaysia based on seven years database (2005-2011). Feed-forward ANN was used as a prediction method. The feed-forward ANN analysis demonstrated that the rotated principal component scores (RPCs) were the best input parameters to predict API. From the 4 RPCs, only 10 (CO, O3, PM10, NO2, CH4, NmHC, THC, wind direction, humidity and ambient temp) out of 12 prediction variables were the most significant parameters to predict API. The results proved that the ANN method can be applied successfully as tools for decision making and problem solving for better atmospheric management.
文摘A multimodal biometric system is applied to recognize individuals for authentication using neural networks. In this paper multimodal biometric algorithm is designed by integrating iris, finger vein, palm print and face biometric traits. Normalized score level fusion approach is applied and optimized, encoded for matching decision. It is a multilevel wavelet, phase based fusion algorithm. This robust multimodal biometric algorithm increases the security level, accuracy, reduces memory size and equal error rate and eliminates unimodal biometric algorithm vulnerabilities.
基金Natural Science Foundation of Jilin Provincial Science and Technology Department (20180101016JC)Science and Technology Development Plan of Jilin Province (20180101054JC).
文摘With the rapid development of the movie industry, it is vital to evaluate and predict a movie’s quality. In this paper, a movie score prediction model is proposed based on the movie plots. Movie data was processed with the word2vec method, and the linear regression model and back propagation neural network algorithm were employed to establish the movie score prediction model. The high-quality classic movie plots of high-scoring movies summed up by big data contributed to a high synthesis of the wonderful content of the film. Experimental results show that it is effective in terms of movie evaluation and prediction, and helpful in understanding people’s preferences for movie plots.
文摘Bank credit is one of the effective ways to meet the financing needs of small and micro enterprises and support the development of inclusive finance. In the credit audit of small and micro enterprises, quantitative risk assessment based on statistical methods plays an increasingly important role. With the development of computer technology and artificial intelligence technology, enterprise general credit evaluation technology is gradually exploring from traditional method to deep learning algorithm. This paper introduces the method of modeling the credit score of small and micro enterprises based on the neural network algorithm, and according to the empirical results, verifies the feasibility of the application of the deep learning algorithm in the credit score of small and micro enterprises.
文摘This study constructs a proposed model to investigate the link between environmental,social,and governance(ESG)disclosures and ESG scores for publicly traded companies in the Borsa Istanbul Sustainability(XUSRD)index.In this context,this study considers 66 companies,examining recently structured ESG disclosures for 2022 that were published for the first time as novel data and applying a multilayer perceptron(MLP)artificial neural network algorithm.The relevant results are fourfold.(1)The MLP algorithm has explanatory power(i.e.,R^(2))of 79% in estimating companies’ESG scores.(2)Common,environment,social,and governance pillars have respective weights of 21.04%,44.87%,30.34%,and 3.74% in total ESG scores.(3)The absolute and relative significance of each ESG reporting principle for companies’ESG scores varies.(4)According to absolute and relative significance,the most effective ESG principle is the common principle,followed by social and environmental principles,whereas governance principles have less significance.Overall,the results demonstrate that applying a linear approach to complete deficient ESG disclosures is inefficient for increasing companies’ESG scores;instead,companies should focus on the ESG principles that have the highest relative significance.The findings of this study contribute to the literature by defining the most significant ESG principles for stimulating the ESG scores of companies in the XUSRD index.
文摘Peer-to-peer (P2P) lending offers an alternative way to access credit. Unlike established lending institutions with proven credit risk management practices, P2P platforms rely on numerous independent variables to evaluate loan applicants’ creditworthiness. This study aims to estimate default probabilities using a mixture-of-experts neural network in P2P lending. The approach involves coupling unsupervised clustering to capture essential data properties with a classification algorithm based on the mixture-of-experts structure. This classic design enhances model capacity without significant computational overhead. The model was tested using P2P data from Lending Club, comparing it to other methods like Logistic Regression, AdaBoost, Gradient Boosting, Decision Tree, Support Vector Machine, and Random Forest. The hybrid model demonstrated superior performance, with a Mean Squared Error reduction of at least 25%.
