To ensure a long-term safety and reliability of electric vehicle and energy storage system,an accurate estimation of the state of health(SOH)for lithium-ion battery is important.In this study,a method for estimating t...To ensure a long-term safety and reliability of electric vehicle and energy storage system,an accurate estimation of the state of health(SOH)for lithium-ion battery is important.In this study,a method for estimating the lithium-ion battery SOH was proposed based on an improved extreme learning machine(ELM).Input weights and hidden layer biases were generated randomly in traditional ELM.To improve the estimation accuracy of ELM,the differential evolution algorithm was used to optimize these parameters in feasible solution spaces.First,incremental capacity curves were obtained by incremental capacity analysis and smoothed by Gaussian filter to extract health interests.Then,the ELM based on differential evolution algorithm(DE-ELM model)was used for a lithium-ion battery SOH estimation.At last,four battery historical aging data sets and one random walk data set were employed to validate the prediction performance of DE-ELM model.Results show that the DE-ELM has a better performance than other studied algorithms in terms of generalization ability.展开更多
As an indispensable task in crop protection,the detection of crop diseases directly impacts the income of farmers.To address the problems of low crop-disease identification precision and detection abilities,a new meth...As an indispensable task in crop protection,the detection of crop diseases directly impacts the income of farmers.To address the problems of low crop-disease identification precision and detection abilities,a new method of detection is proposed based on improved genetic algorithm and extreme learning machine.Taking five different typical diseases with common crops as the objects,this method first preprocesses the images of crops and selects the optimal features for fusion.Then,it builds a model of crop disease identification for extreme learning machine,introduces the hill-climbing algorithm to improve the traditional genetic algorithm,optimizes the initial weights and thresholds of the machine,and acquires the approximately optimal solution.And finally,a data set of crop diseases is used for verification,demonstrating that,compared with several other common machine learning methods,this method can effectively improve the crop-disease identification precision and detection abilities and provide a basis for the identification of other crop diseases.展开更多
Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of ...Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of data from unstructured literature.To overcome this,we introduce a transformative“LLM to Agent”framework that synergistically integrates Large Language Models(LLMs)for automated data curation with Machine Learning(ML)for predictive design.We automatically constructed a comprehensive database of 809 MgH_(2)catalysts(6555 data rows)with high fidelity and an~40-fold acceleration over manual methods.The resulting ML models achieved high accuracy(average R^(2)>0.91)in predicting dehydrogenation temperature and activation energy,subsequently guiding a Genetic Algorithm(GA)in an exploratory inverse design that autonomously uncovered key design principles for high-performance catalysts.Encouragingly,a strong alignment was found between these AI-discovered principles and the design strategies of recently reported,state-of-the-art experimental systems,providing substantial evidence for the validity of our approach.The framework culminates in Cat-Advisor,a novel,domain-adapted multi-agent system.Cat-Advisor translates ML predictions and retrieval-augmented knowledge into actionable design guidance,demonstrating capabilities that surpass those of general-purpose LLMs in this specialized domain.This work delivers a practical AI toolkit for accelerated materials discovery and advances the emerging Agent-based paradigm for designing next-generation energy technologies.展开更多
To curb the worsening tropospheric ozone(O_(3))pollution problem in China,a rapid and accurate identification of O_(3)-precursor sensitivity(OPS)is a crucial prerequisite for formulating effective contingency O_(3) po...To curb the worsening tropospheric ozone(O_(3))pollution problem in China,a rapid and accurate identification of O_(3)-precursor sensitivity(OPS)is a crucial prerequisite for formulating effective contingency O_(3) pollution control strategies.However,currently widely-used methods,such as statistical models and numerical models,exhibit inherent limitations in identifying OPS in a timely and accurate manner.In this study,we developed a novel approach to identify OPS based on eXtreme Gradient Boosting model,Shapley additive explanation(SHAP)al-gorithm,and volatile organic compound(VOC)photochemical decay adjustment,using the meteorology and speciated pollutant monitoring data as the input.By comparing the difference in SHAP values between base sce-nario and precursor reduction scenario for nitrogen oxides(NO_(x))and VOCs,OPS was divided into NO_(x)-limited,VOCs-limited and transition regime.Using the long-lasting O_(3) pollution episode in the autumn of 2022 at the Guangdong-Hong Kong-Macao Greater Bay Area(GBA)as an example,we demonstrated large spatiotemporal heterogeneities of OPS over the GBA,which were generally shifted from NO_(x)-limited to VOCs-limited from September to October and more inclined to be VOCs-limited at the central and NO_(x)-limited in the peripheral areas.This study developed an innovative OPS identification method by comparing the difference in SHAP value before and after precursor emission reduction.Our method enables the accurate identification of OPS in the time scale of seconds,thereby providing a state-of-the-art tool for the rapid guidance of spatial-specific O_(3) control strategies.展开更多
As one of the most serious geological disasters in deep underground engineering,rockburst has caused a large number of casualties.However,because of the complex relationship between the inducing factors and rockburst ...As one of the most serious geological disasters in deep underground engineering,rockburst has caused a large number of casualties.However,because of the complex relationship between the inducing factors and rockburst intensity,the problem of rockburst intensity prediction has not been well solved until now.In this study,we collect 292 sets of rockburst data including eight parameters,such as the maximum tangential stress of the surrounding rock σ_(θ),the uniaxial compressive strength of the rockσc,the uniaxial tensile strength of the rock σ_(t),and the strain energy storage index W_(et),etc.from more than 20 underground projects as training sets and establish two new rockburst prediction models based on the kernel extreme learning machine(KELM)combined with the genetic algorithm(KELM-GA)and cross-entropy method(KELM-CEM).To further verify the effect of the two models,ten sets of rockburst data from Shuangjiangkou Hydropower Station are selected for analysis and the results show that new models are more accurate compared with five traditional empirical criteria,especially the model based on KELM-CEM which has the accuracy rate of 90%.Meanwhile,the results of 10 consecutive runs of the model based on KELM-CEM are almost the same,meaning that the model has good stability and reliability for engineering applications.展开更多
Bimetallic nanoparticles(AmBn)usually exhibit rich catalytic chemistry and have drawn tremendous attention in heterogeneous catalysis.However,challenged by the huge configuration space,the understanding toward their c...Bimetallic nanoparticles(AmBn)usually exhibit rich catalytic chemistry and have drawn tremendous attention in heterogeneous catalysis.However,challenged by the huge configuration space,the understanding toward their composition and distribution of A/B element is known little at the atomic level,which hinders the rational synthesis.Herein,we develop an on-the-fly training strategy combing the machine learning model(SchNet)with the genetic algorithm(GA)search technique,which achieve the fast and accurate energy prediction of complex bimetallic clusters at the DFT level.Taking the 38-atom PtmAu38-mnanoparticle as example,the element distribution identification problem and the stability trend as a function of Pt/Au composition is quantitatively re solved.Specifically,results show that on the Pt-rich cluster Au atoms prefer to occupy the low-coordinated surface corner sites and form patch-like surface segregation patte rns,while for the Au-rich ones Pt atoms tend to site in the co re region and form the co re-shell(Pt@Au)configuration.The thermodynamically most stable PtmAu38-mcluster is Pt6 Au32,with all the core-region sites occupied by Pt,rationalized by the stronger Pt-Pt bond in comparison with Pt-Au and Au-Au bonds.This work exemplifies the potent application of rapid global sea rch enabled by machine learning in exploring the high-dimensional configuration space of bimetallic nanocatalysts.展开更多
In order to improve the accuracy and reduce the training and testing time in image classification algorithm, a novel image classification scheme based on extreme learning machine(ELM) and linear spatial pyramid matchi...In order to improve the accuracy and reduce the training and testing time in image classification algorithm, a novel image classification scheme based on extreme learning machine(ELM) and linear spatial pyramid matching using sparse coding(Sc SPM) for image classification is proposed. A new structure based on two layer extreme learning machine instead of the original linear SVM classifier is constructed. Firstly, the Sc SPM algorithm is performed to extract features of the multi-scale image blocks, and then each layer feature vector is connected to an ELM. Finally, the mapping features are connected together, and as the input of one ELM based on radial basis kernel function. With experimental evaluations on the well-known dataset benchmarks, the results demonstrate that the proposed algorithm has better performance not only in reducing the training time, but also in improving the accuracy of classification.展开更多
Background:Diabetic nephropathy(DN)is the most common complication of type 2 diabetes mellitus and the main cause of end-stage renal disease worldwide.Diagnostic biomarkers may allow early diagnosis and treatment of D...Background:Diabetic nephropathy(DN)is the most common complication of type 2 diabetes mellitus and the main cause of end-stage renal disease worldwide.Diagnostic biomarkers may allow early diagnosis and treatment of DN to reduce the prevalence and delay the development of DN.Kidney biopsy is the gold standard for diagnosing DN;however,its invasive character is its primary limitation.The machine learning approach provides a non-invasive and specific criterion for diagnosing DN,although traditional machine learning algorithms need to be improved to enhance diagnostic performance.Methods:We applied high-throughput RNA sequencing to obtain the genes related to DN tubular tissues and normal tubular tissues of mice.Then machine learning algorithms,random forest,LASSO logistic regression,and principal component analysis were used to identify key genes(CES1G,CYP4A14,NDUFA4,ABCC4,ACE).Then,the genetic algorithm-optimized backpropagation neural network(GA-BPNN)was used to improve the DN diagnostic model.Results:The AUC value of the GA-BPNN model in the training dataset was 0.83,and the AUC value of the model in the validation dataset was 0.81,while the AUC values of the SVM model in the training dataset and external validation dataset were 0.756 and 0.650,respectively.Thus,this GA-BPNN gave better values than the traditional SVM model.This diagnosis model may aim for personalized diagnosis and treatment of patients with DN.Immunohistochemical staining further confirmed that the tissue and cell expression of NADH dehydrogenase(ubiquinone)1 alpha subcomplex,4-like 2(NDUFA4L2)in tubular tissue in DN mice were decreased.Conclusion:The GA-BPNN model has better accuracy than the traditional SVM model and may provide an effective tool for diagnosing DN.展开更多
In propylene polymerization(PP) process, the melt index(MI) is one of the most important quality variables for determining different brands of products and different grades of product quality. Accurate prediction of M...In propylene polymerization(PP) process, the melt index(MI) is one of the most important quality variables for determining different brands of products and different grades of product quality. Accurate prediction of MI is essential for efficient and professional monitoring and control of practical PP processes. This paper presents a novel soft sensor based on extreme learning machine(ELM) and modified gravitational search algorithm(MGSA) to estimate MI from real PP process variables, where the MGSA algorithm is developed to find the best parameters of input weights and hidden biases for ELM. As the comparative basis, the models of ELM, APSO-ELM and GSAELM are also developed respectively. Based on the data from a real PP production plant, a detailed comparison of the models is carried out. The research results show the accuracy and universality of the proposed model and it can be a powerful tool for online MI prediction.展开更多
Compression index Ccis an essential parameter in geotechnical design for which the effectiveness of correlation is still a challenge.This paper suggests a novel modelling approach using machine learning(ML)technique.T...Compression index Ccis an essential parameter in geotechnical design for which the effectiveness of correlation is still a challenge.This paper suggests a novel modelling approach using machine learning(ML)technique.The performance of five commonly used machine learning(ML)algorithms,i.e.back-propagation neural network(BPNN),extreme learning machine(ELM),support vector machine(SVM),random forest(RF)and evolutionary polynomial regression(EPR)in predicting Cc is comprehensively investigated.A database with a total number of 311 datasets including three input variables,i.e.initial void ratio e0,liquid limit water content wL,plasticity index Ip,and one output variable Cc is first established.Genetic algorithm(GA)is used to optimize the hyper-parameters in five ML algorithms,and the average prediction error for the 10-fold cross-validation(CV)sets is set as thefitness function in the GA for enhancing the robustness of ML models.The results indicate that ML models outperform empirical prediction formulations with lower prediction error.RF yields the lowest error followed by BPNN,ELM,EPR and SVM.If the ranges of input variables in the database are large enough,BPNN and RF models are recommended to predict Cc.Furthermore,if the distribution of input variables is continuous,RF model is the best one.Otherwise,EPR model is recommended if the ranges of input variables are small.The predicted correlations between input and output variables using five ML models show great agreement with the physical explanation.展开更多
Real-time and reliable measurements of the effluent quality are essential to improve operating efficiency and reduce energy consumption for the wastewater treatment process.Due to the low accuracy and unstable perform...Real-time and reliable measurements of the effluent quality are essential to improve operating efficiency and reduce energy consumption for the wastewater treatment process.Due to the low accuracy and unstable performance of the traditional effluent quality measurements,we propose a selective ensemble extreme learning machine modeling method to enhance the effluent quality predictions.Extreme learning machine algorithm is inserted into a selective ensemble frame as the component model since it runs much faster and provides better generalization performance than other popular learning algorithms.Ensemble extreme learning machine models overcome variations in different trials of simulations for single model.Selective ensemble based on genetic algorithm is used to further exclude some bad components from all the available ensembles in order to reduce the computation complexity and improve the generalization performance.The proposed method is verified with the data from an industrial wastewater treatment plant,located in Shenyang,China.Experimental results show that the proposed method has relatively stronger generalization and higher accuracy than partial least square,neural network partial least square,single extreme learning machine and ensemble extreme learning machine model.展开更多
A hybrid feature selection and classification strategy was proposed based on the simulated annealing genetic algonthrn and multiple instance learning （MIL）. The band selection method was proposed from subspace decom...A hybrid feature selection and classification strategy was proposed based on the simulated annealing genetic algonthrn and multiple instance learning （MIL）. The band selection method was proposed from subspace decomposition, which combines the simulated annealing algorithm with the genetic algorithm in choosing different cross-over and mutation probabilities, as well as mutation individuals. Then MIL was combined with image segmentation, clustering and support vector machine algorithms to classify hyperspectral image. The experimental results show that this proposed method can get high classification accuracy of 93.13% at small training samples and the weaknesses of the conventional methods are overcome.展开更多
Traditional geostatistical estimation techniques have been used predominantly by the mining industry for ore reserve estimation. Determination of mineral reserve has posed considerable challenge to mining engineers du...Traditional geostatistical estimation techniques have been used predominantly by the mining industry for ore reserve estimation. Determination of mineral reserve has posed considerable challenge to mining engineers due to the geological complexities of ore body formation. Extensive research over the years has resulted in the development of several state-of-the-art methods for predictive spatial mapping, which could be used for ore reserve estimation;and recent advances in the use of machine learning algorithms (MLA) have provided a new approach for solving the prob-lem of ore reserve estimation. The focus of the present study was on the use of two MLA for estimating ore reserve: namely, neural networks (NN) and support vector machines (SVM). Application of MLA and the various issues involved with using them for reserve estimation have been elaborated with the help of a complex drill-hole dataset that exhibits the typical properties of sparseness and impreciseness that might be associated with a mining dataset. To investigate the accuracy and applicability of MLA for ore reserve estimation, the generalization ability of NN and SVM was compared with the geostatistical ordinary kriging (OK) method.展开更多
Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machin...Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machine(RELM)is one of them.However,the regularization parameter determination of RELM consumes computational resources,which makes it unsuitable in the field of aeroengine performance parameter prediction with a large amount of data.This paper uses the forward and backward segmentation(FBS)algorithms to improve the RELM performance,and introduces an adaptive step size determination method and an improved solution mechanism to obtain a new machine learning algorithm.While maintaining good generalization,the new algorithm is not sensitive to regularization parameters,which greatly saves computing resources.The experimental results on the public data sets prove the above conclusions.Finally,the new algorithm is applied to the prediction of aero-engine performance parameters,and the excellent prediction performance is achieved.展开更多
An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant cont...An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant contribution to cyber-security techniques.Multiple techniques have been in use but major concern in their implementation is variation in their detection performance.The performance of IDS lies in the accurate detection of attacks,and this accuracy can be raised by improving the recognition rate and significant reduction in the false alarms rate.To overcome this problem many researchers have used different machine learning techniques.These techniques have limitations and do not efficiently perform on huge and complex data about systems and networks.This work focused on ELM(Extreme Learning Machine)technique due to its good capabilities in classification problems and dealing with huge data.The ELM has different activation functions,but the problem is to find out which function is more suitable and performs well in IDS.This work investigates this problem.Here,Well-known activation functions like:sine,sigmoid and radial basis are explored,investigated and applied to measure their performance on the GA(Genetic Algorithm)features subset and with full features set.The NSL-KDD dataset is used as a benchmark.The empirical results are analyzed,addressed and compared among different activation functions of the ELM.The results show that the radial basis and sine functions perform better on GA feature set than the full feature set while the performance of the sigmoid function is almost equal on both features sets.So,the proposal of GA based feature selection reduced 21 features out of 41 that brought up to 98%accuracy and enhanced overall efficiency of extreme learning machine in intrusion detection.展开更多
Metallic alloys for a given application are usually designed to achieve the desired properties by devising experimentsbased on experience, thermodynamic and kinetic principles, and various modeling and simulation exer...Metallic alloys for a given application are usually designed to achieve the desired properties by devising experimentsbased on experience, thermodynamic and kinetic principles, and various modeling and simulation exercises.However, the influence of process parameters and material properties is often non-linear and non-colligative. Inrecent years, machine learning (ML) has emerged as a promising tool to dealwith the complex interrelation betweencomposition, properties, and process parameters to facilitate accelerated discovery and development of new alloysand functionalities. In this study, we adopt an ML-based approach, coupled with genetic algorithm (GA) principles,to design novel copper alloys for achieving seemingly contradictory targets of high strength and high electricalconductivity. Initially, we establish a correlation between the alloy composition (binary to multi-component) andthe target properties, namely, electrical conductivity and mechanical strength. Catboost, an ML model coupledwith GA, was used for this task. The accuracy of the model was above 93.5%. Next, for obtaining the optimizedcompositions the outputs fromthe initial model were refined by combining the concepts of data augmentation andPareto front. Finally, the ultimate objective of predicting the target composition that would deliver the desired rangeof properties was achieved by developing an advancedMLmodel through data segregation and data augmentation.To examine the reliability of this model, results were rigorously compared and verified using several independentdata reported in the literature. This comparison substantiates that the results predicted by our model regarding thevariation of conductivity and evolution ofmicrostructure and mechanical properties with composition are in goodagreement with the reports published in the literature.展开更多
Extreme Learning Machine(ELM)is popular in batch learning,sequential learning,and progressive learning,due to its speed,easy integration,and generalization ability.While,Traditional ELM cannot train massive data rapid...Extreme Learning Machine(ELM)is popular in batch learning,sequential learning,and progressive learning,due to its speed,easy integration,and generalization ability.While,Traditional ELM cannot train massive data rapidly and efficiently due to its memory residence,high time and space complexity.In ELM,the hidden layer typically necessitates a huge number of nodes.Furthermore,there is no certainty that the arrangement of weights and biases within the hidden layer is optimal.To solve this problem,the traditional ELM has been hybridized with swarm intelligence optimization techniques.This paper displays five proposed hybrid Algorithms“Salp Swarm Algorithm(SSA-ELM),Grasshopper Algorithm(GOA-ELM),Grey Wolf Algorithm(GWO-ELM),Whale optimizationAlgorithm(WOA-ELM)andMoth Flame Optimization(MFO-ELM)”.These five optimizers are hybridized with standard ELM methodology for resolving the tumor type classification using gene expression data.The proposed models applied to the predication of electricity loading data,that describes the energy use of a single residence over a fouryear period.In the hidden layer,Swarm algorithms are used to pick a smaller number of nodes to speed up the execution of ELM.The best weights and preferences were calculated by these algorithms for the hidden layer.Experimental results demonstrated that the proposed MFO-ELM achieved 98.13%accuracy and this is the highest model in accuracy in tumor type classification gene expression data.While in predication,the proposed GOA-ELM achieved 0.397which is least RMSE compared to the other models.展开更多
Support vector machines （SVMs） have been introduced as effective methods for solving classification problems. However, due to some limitations in practical applications, their generalization performance is sometimes...Support vector machines （SVMs） have been introduced as effective methods for solving classification problems. However, due to some limitations in practical applications, their generalization performance is sometimes far from the expected level. Therefore, it is meaningful to study SVM ensemble learning. In this paper, a novel genetic algorithm based ensemble learning method, namely Direct Genetic Ensemble （DGE）, is proposed. DGE adopts the predictive accuracy of ensemble as the fitness function and searches a good ensemble from the ensemble space. In essence, DGE is also a selective ensemble learning method because the base classifiers of the ensemble are selected according to the solution of genetic algorithm. In comparison with other ensemble learning methods, DGE works on a higher level and is more direct. Different strategies of constructing diverse base classifiers can be utilized in DGE. Experimental results show that SVM ensembles constructed by DGE can achieve better performance than single SVMs, hagged and boosted SVM ensembles. In addition, some valuable conclusions are obtained.展开更多
Breast cancer diagnosis through mammography is a pivotal application within medical image-based diagnostics,integral for early detection and effective treatment.While deep learning has significantly advanced the analy...Breast cancer diagnosis through mammography is a pivotal application within medical image-based diagnostics,integral for early detection and effective treatment.While deep learning has significantly advanced the analysis of mammographic images,challenges such as low contrast,image noise,and the high dimensionality of features often degrade model performance.Addressing these challenges,our study introduces a novel method integrating Genetic Algorithms(GA)with pre-trained Convolutional Neural Network(CNN)models to enhance feature selection and classification accuracy.Our approach involves a systematic process:first,we employ widely-used CNN architectures(VGG16,VGG19,MobileNet,and DenseNet)to extract a broad range of features from the Medical Image Analysis Society(MIAS)mammography dataset.Subsequently,a GA optimizes these features by selecting the most relevant and least redundant,aiming to overcome the typical pitfalls of high dimensionality.The selected features are then utilized to train several classifiers,including Linear and Polynomial Support Vector Machines(SVMs),K-Nearest Neighbors,Decision Trees,and Random Forests,enabling a robust evaluation of the method’s effectiveness across varied learning algorithms.Our extensive experimental evaluation demonstrates that the integration of MobileNet and GA significantly improves classification accuracy,from 83.33%to 89.58%,underscoring the method’s efficacy.By detailing these steps,we highlight the innovation of our approach which not only addresses key issues in breast cancer imaging analysis but also offers a scalable solution potentially applicable to other domains within medical imaging.展开更多
Neural networks(NNs),as one of the most robust and efficient machine learning methods,have been commonly used in solving several problems.However,choosing proper hyperparameters(e.g.the numbers of layers and neurons i...Neural networks(NNs),as one of the most robust and efficient machine learning methods,have been commonly used in solving several problems.However,choosing proper hyperparameters(e.g.the numbers of layers and neurons in each layer)has a significant influence on the accuracy of these methods.Therefore,a considerable number of studies have been carried out to optimize the NN hyperpaxameters.In this study,the genetic algorithm is applied to NN to find the optimal hyperpaxameters.Thus,the deep energy method,which contains a deep neural network,is applied first on a Timoshenko beam and a plate with a hole.Subsequently,the numbers of hidden layers,integration points,and neurons in each layer are optimized to reach the highest accuracy to predict the stress distribution through these structures.Thus,applying the proper optimization method on NN leads to significant increase in the NN prediction accuracy after conducting the optimization in various examples.展开更多
文摘To ensure a long-term safety and reliability of electric vehicle and energy storage system,an accurate estimation of the state of health(SOH)for lithium-ion battery is important.In this study,a method for estimating the lithium-ion battery SOH was proposed based on an improved extreme learning machine(ELM).Input weights and hidden layer biases were generated randomly in traditional ELM.To improve the estimation accuracy of ELM,the differential evolution algorithm was used to optimize these parameters in feasible solution spaces.First,incremental capacity curves were obtained by incremental capacity analysis and smoothed by Gaussian filter to extract health interests.Then,the ELM based on differential evolution algorithm(DE-ELM model)was used for a lithium-ion battery SOH estimation.At last,four battery historical aging data sets and one random walk data set were employed to validate the prediction performance of DE-ELM model.Results show that the DE-ELM has a better performance than other studied algorithms in terms of generalization ability.
基金This paper is supported by the National Youth Natural Science Foundation of China(61802208)the National Natural Science Foundation of China(61572261)+4 种基金the Natural Science Foundation of Anhui(1908085MF207 and 1908085QE217)the Excellent Youth Talent Support Foundation of Anhui(gxyqZD2019097)the Postdoctoral Foundation of Jiangsu(2018K009B)the Higher Education Quality Project of Anhui(2019sjjd81,2018mooc059,2018kfk009,2018sxzx38 and 2018FXJT02)the Fuyang Normal University Doctoral Startup Foundation and Fuyang Government Research Foundation(2017KYQD0008 and XDHXTD201703).
文摘As an indispensable task in crop protection,the detection of crop diseases directly impacts the income of farmers.To address the problems of low crop-disease identification precision and detection abilities,a new method of detection is proposed based on improved genetic algorithm and extreme learning machine.Taking five different typical diseases with common crops as the objects,this method first preprocesses the images of crops and selects the optimal features for fusion.Then,it builds a model of crop disease identification for extreme learning machine,introduces the hill-climbing algorithm to improve the traditional genetic algorithm,optimizes the initial weights and thresholds of the machine,and acquires the approximately optimal solution.And finally,a data set of crop diseases is used for verification,demonstrating that,compared with several other common machine learning methods,this method can effectively improve the crop-disease identification precision and detection abilities and provide a basis for the identification of other crop diseases.
基金supported by the Natural Science Foundation of Hebei Province(E2023502006)Fundamental Research Fund for the Central Universities(2025MS131).
文摘Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of data from unstructured literature.To overcome this,we introduce a transformative“LLM to Agent”framework that synergistically integrates Large Language Models(LLMs)for automated data curation with Machine Learning(ML)for predictive design.We automatically constructed a comprehensive database of 809 MgH_(2)catalysts(6555 data rows)with high fidelity and an~40-fold acceleration over manual methods.The resulting ML models achieved high accuracy(average R^(2)>0.91)in predicting dehydrogenation temperature and activation energy,subsequently guiding a Genetic Algorithm(GA)in an exploratory inverse design that autonomously uncovered key design principles for high-performance catalysts.Encouragingly,a strong alignment was found between these AI-discovered principles and the design strategies of recently reported,state-of-the-art experimental systems,providing substantial evidence for the validity of our approach.The framework culminates in Cat-Advisor,a novel,domain-adapted multi-agent system.Cat-Advisor translates ML predictions and retrieval-augmented knowledge into actionable design guidance,demonstrating capabilities that surpass those of general-purpose LLMs in this specialized domain.This work delivers a practical AI toolkit for accelerated materials discovery and advances the emerging Agent-based paradigm for designing next-generation energy technologies.
基金supported by the Key-Area Research and Development Program of Guangdong Province(No.2020B1111360003)the National Natural Science Foundation of China(Nos.42465008 and 42105164)+2 种基金Yunnan Science and Technology Department Project(No.202501AT070239)Yunnan Science and Technology Department Youth Project(No.202401AU070202)Xianyang Rapid Response Decision Support Project for Ozone(No.YZ2024-ZB019).
文摘To curb the worsening tropospheric ozone(O_(3))pollution problem in China,a rapid and accurate identification of O_(3)-precursor sensitivity(OPS)is a crucial prerequisite for formulating effective contingency O_(3) pollution control strategies.However,currently widely-used methods,such as statistical models and numerical models,exhibit inherent limitations in identifying OPS in a timely and accurate manner.In this study,we developed a novel approach to identify OPS based on eXtreme Gradient Boosting model,Shapley additive explanation(SHAP)al-gorithm,and volatile organic compound(VOC)photochemical decay adjustment,using the meteorology and speciated pollutant monitoring data as the input.By comparing the difference in SHAP values between base sce-nario and precursor reduction scenario for nitrogen oxides(NO_(x))and VOCs,OPS was divided into NO_(x)-limited,VOCs-limited and transition regime.Using the long-lasting O_(3) pollution episode in the autumn of 2022 at the Guangdong-Hong Kong-Macao Greater Bay Area(GBA)as an example,we demonstrated large spatiotemporal heterogeneities of OPS over the GBA,which were generally shifted from NO_(x)-limited to VOCs-limited from September to October and more inclined to be VOCs-limited at the central and NO_(x)-limited in the peripheral areas.This study developed an innovative OPS identification method by comparing the difference in SHAP value before and after precursor emission reduction.Our method enables the accurate identification of OPS in the time scale of seconds,thereby providing a state-of-the-art tool for the rapid guidance of spatial-specific O_(3) control strategies.
基金funded by National Natural Science Foundation of China(Grants Nos.41825018 and 42141009)the Second Tibetan Plateau Scientific Expedition and Research Program(Grants No.2019QZKK0904)。
文摘As one of the most serious geological disasters in deep underground engineering,rockburst has caused a large number of casualties.However,because of the complex relationship between the inducing factors and rockburst intensity,the problem of rockburst intensity prediction has not been well solved until now.In this study,we collect 292 sets of rockburst data including eight parameters,such as the maximum tangential stress of the surrounding rock σ_(θ),the uniaxial compressive strength of the rockσc,the uniaxial tensile strength of the rock σ_(t),and the strain energy storage index W_(et),etc.from more than 20 underground projects as training sets and establish two new rockburst prediction models based on the kernel extreme learning machine(KELM)combined with the genetic algorithm(KELM-GA)and cross-entropy method(KELM-CEM).To further verify the effect of the two models,ten sets of rockburst data from Shuangjiangkou Hydropower Station are selected for analysis and the results show that new models are more accurate compared with five traditional empirical criteria,especially the model based on KELM-CEM which has the accuracy rate of 90%.Meanwhile,the results of 10 consecutive runs of the model based on KELM-CEM are almost the same,meaning that the model has good stability and reliability for engineering applications.
基金supported by National Key R&D Program of China(No.2018YFA0208602)NSFC(Nos.21622305,21873028,21703067)+1 种基金National Ten Thousand Talent Program for Young Top-notch Talents in ChinaShanghai ShuGuang project(No.17SG30)。
文摘Bimetallic nanoparticles(AmBn)usually exhibit rich catalytic chemistry and have drawn tremendous attention in heterogeneous catalysis.However,challenged by the huge configuration space,the understanding toward their composition and distribution of A/B element is known little at the atomic level,which hinders the rational synthesis.Herein,we develop an on-the-fly training strategy combing the machine learning model(SchNet)with the genetic algorithm(GA)search technique,which achieve the fast and accurate energy prediction of complex bimetallic clusters at the DFT level.Taking the 38-atom PtmAu38-mnanoparticle as example,the element distribution identification problem and the stability trend as a function of Pt/Au composition is quantitatively re solved.Specifically,results show that on the Pt-rich cluster Au atoms prefer to occupy the low-coordinated surface corner sites and form patch-like surface segregation patte rns,while for the Au-rich ones Pt atoms tend to site in the co re region and form the co re-shell(Pt@Au)configuration.The thermodynamically most stable PtmAu38-mcluster is Pt6 Au32,with all the core-region sites occupied by Pt,rationalized by the stronger Pt-Pt bond in comparison with Pt-Au and Au-Au bonds.This work exemplifies the potent application of rapid global sea rch enabled by machine learning in exploring the high-dimensional configuration space of bimetallic nanocatalysts.
基金supported by the National Natural Science Foundation of China under Grant No. 61503424the Research Project by The State Ethnic Affairs Commission under Grant No. 14ZYZ017+1 种基金the Jiangsu Future Networks Innovation Institute-Prospective Research Project on Future Networks under Grant No. BY2013095-2-14the first-class discipline construction transitional funds of Minzu University of China
文摘In order to improve the accuracy and reduce the training and testing time in image classification algorithm, a novel image classification scheme based on extreme learning machine(ELM) and linear spatial pyramid matching using sparse coding(Sc SPM) for image classification is proposed. A new structure based on two layer extreme learning machine instead of the original linear SVM classifier is constructed. Firstly, the Sc SPM algorithm is performed to extract features of the multi-scale image blocks, and then each layer feature vector is connected to an ELM. Finally, the mapping features are connected together, and as the input of one ELM based on radial basis kernel function. With experimental evaluations on the well-known dataset benchmarks, the results demonstrate that the proposed algorithm has better performance not only in reducing the training time, but also in improving the accuracy of classification.
基金the National Natural Science Foundation of China(Grant Number:81970631 to W.L.).
文摘Background:Diabetic nephropathy(DN)is the most common complication of type 2 diabetes mellitus and the main cause of end-stage renal disease worldwide.Diagnostic biomarkers may allow early diagnosis and treatment of DN to reduce the prevalence and delay the development of DN.Kidney biopsy is the gold standard for diagnosing DN;however,its invasive character is its primary limitation.The machine learning approach provides a non-invasive and specific criterion for diagnosing DN,although traditional machine learning algorithms need to be improved to enhance diagnostic performance.Methods:We applied high-throughput RNA sequencing to obtain the genes related to DN tubular tissues and normal tubular tissues of mice.Then machine learning algorithms,random forest,LASSO logistic regression,and principal component analysis were used to identify key genes(CES1G,CYP4A14,NDUFA4,ABCC4,ACE).Then,the genetic algorithm-optimized backpropagation neural network(GA-BPNN)was used to improve the DN diagnostic model.Results:The AUC value of the GA-BPNN model in the training dataset was 0.83,and the AUC value of the model in the validation dataset was 0.81,while the AUC values of the SVM model in the training dataset and external validation dataset were 0.756 and 0.650,respectively.Thus,this GA-BPNN gave better values than the traditional SVM model.This diagnosis model may aim for personalized diagnosis and treatment of patients with DN.Immunohistochemical staining further confirmed that the tissue and cell expression of NADH dehydrogenase(ubiquinone)1 alpha subcomplex,4-like 2(NDUFA4L2)in tubular tissue in DN mice were decreased.Conclusion:The GA-BPNN model has better accuracy than the traditional SVM model and may provide an effective tool for diagnosing DN.
基金Supported by the Major Program of National Natural Science Foundation of China(61590921)the Natural Science Foundation of Zhejiang Province(Y16B040003)+1 种基金Shanghai Aerospace Science and Technology Innovation Fund(E11501)Aerospace Science and Technology Innovation Fund of China,Aerospace Science and Technology Corporation(E11601)
文摘In propylene polymerization(PP) process, the melt index(MI) is one of the most important quality variables for determining different brands of products and different grades of product quality. Accurate prediction of MI is essential for efficient and professional monitoring and control of practical PP processes. This paper presents a novel soft sensor based on extreme learning machine(ELM) and modified gravitational search algorithm(MGSA) to estimate MI from real PP process variables, where the MGSA algorithm is developed to find the best parameters of input weights and hidden biases for ELM. As the comparative basis, the models of ELM, APSO-ELM and GSAELM are also developed respectively. Based on the data from a real PP production plant, a detailed comparison of the models is carried out. The research results show the accuracy and universality of the proposed model and it can be a powerful tool for online MI prediction.
基金financial support provided by the RIF project(Grant No.PolyU R5037-18F)from the Research Grants Council(RGC)of Hong Kong is gratefully acknowledged。
文摘Compression index Ccis an essential parameter in geotechnical design for which the effectiveness of correlation is still a challenge.This paper suggests a novel modelling approach using machine learning(ML)technique.The performance of five commonly used machine learning(ML)algorithms,i.e.back-propagation neural network(BPNN),extreme learning machine(ELM),support vector machine(SVM),random forest(RF)and evolutionary polynomial regression(EPR)in predicting Cc is comprehensively investigated.A database with a total number of 311 datasets including three input variables,i.e.initial void ratio e0,liquid limit water content wL,plasticity index Ip,and one output variable Cc is first established.Genetic algorithm(GA)is used to optimize the hyper-parameters in five ML algorithms,and the average prediction error for the 10-fold cross-validation(CV)sets is set as thefitness function in the GA for enhancing the robustness of ML models.The results indicate that ML models outperform empirical prediction formulations with lower prediction error.RF yields the lowest error followed by BPNN,ELM,EPR and SVM.If the ranges of input variables in the database are large enough,BPNN and RF models are recommended to predict Cc.Furthermore,if the distribution of input variables is continuous,RF model is the best one.Otherwise,EPR model is recommended if the ranges of input variables are small.The predicted correlations between input and output variables using five ML models show great agreement with the physical explanation.
基金supported by National Natural Science Foundation of China(Nos.61203102 and 60874057)Postdoctoral Science Foundation of China(No.20100471464)
文摘Real-time and reliable measurements of the effluent quality are essential to improve operating efficiency and reduce energy consumption for the wastewater treatment process.Due to the low accuracy and unstable performance of the traditional effluent quality measurements,we propose a selective ensemble extreme learning machine modeling method to enhance the effluent quality predictions.Extreme learning machine algorithm is inserted into a selective ensemble frame as the component model since it runs much faster and provides better generalization performance than other popular learning algorithms.Ensemble extreme learning machine models overcome variations in different trials of simulations for single model.Selective ensemble based on genetic algorithm is used to further exclude some bad components from all the available ensembles in order to reduce the computation complexity and improve the generalization performance.The proposed method is verified with the data from an industrial wastewater treatment plant,located in Shenyang,China.Experimental results show that the proposed method has relatively stronger generalization and higher accuracy than partial least square,neural network partial least square,single extreme learning machine and ensemble extreme learning machine model.
文摘A hybrid feature selection and classification strategy was proposed based on the simulated annealing genetic algonthrn and multiple instance learning （MIL）. The band selection method was proposed from subspace decomposition, which combines the simulated annealing algorithm with the genetic algorithm in choosing different cross-over and mutation probabilities, as well as mutation individuals. Then MIL was combined with image segmentation, clustering and support vector machine algorithms to classify hyperspectral image. The experimental results show that this proposed method can get high classification accuracy of 93.13% at small training samples and the weaknesses of the conventional methods are overcome.
文摘Traditional geostatistical estimation techniques have been used predominantly by the mining industry for ore reserve estimation. Determination of mineral reserve has posed considerable challenge to mining engineers due to the geological complexities of ore body formation. Extensive research over the years has resulted in the development of several state-of-the-art methods for predictive spatial mapping, which could be used for ore reserve estimation;and recent advances in the use of machine learning algorithms (MLA) have provided a new approach for solving the prob-lem of ore reserve estimation. The focus of the present study was on the use of two MLA for estimating ore reserve: namely, neural networks (NN) and support vector machines (SVM). Application of MLA and the various issues involved with using them for reserve estimation have been elaborated with the help of a complex drill-hole dataset that exhibits the typical properties of sparseness and impreciseness that might be associated with a mining dataset. To investigate the accuracy and applicability of MLA for ore reserve estimation, the generalization ability of NN and SVM was compared with the geostatistical ordinary kriging (OK) method.
文摘Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machine(RELM)is one of them.However,the regularization parameter determination of RELM consumes computational resources,which makes it unsuitable in the field of aeroengine performance parameter prediction with a large amount of data.This paper uses the forward and backward segmentation(FBS)algorithms to improve the RELM performance,and introduces an adaptive step size determination method and an improved solution mechanism to obtain a new machine learning algorithm.While maintaining good generalization,the new algorithm is not sensitive to regularization parameters,which greatly saves computing resources.The experimental results on the public data sets prove the above conclusions.Finally,the new algorithm is applied to the prediction of aero-engine performance parameters,and the excellent prediction performance is achieved.
基金This project was funded by the Deanship of Scientific Research(DSR)at King Abdulaziz University,Jeddah under grant no.G:656-611-1439The authors,therefore,acknowledge with thanks DSR for technical and financial support.
文摘An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant contribution to cyber-security techniques.Multiple techniques have been in use but major concern in their implementation is variation in their detection performance.The performance of IDS lies in the accurate detection of attacks,and this accuracy can be raised by improving the recognition rate and significant reduction in the false alarms rate.To overcome this problem many researchers have used different machine learning techniques.These techniques have limitations and do not efficiently perform on huge and complex data about systems and networks.This work focused on ELM(Extreme Learning Machine)technique due to its good capabilities in classification problems and dealing with huge data.The ELM has different activation functions,but the problem is to find out which function is more suitable and performs well in IDS.This work investigates this problem.Here,Well-known activation functions like:sine,sigmoid and radial basis are explored,investigated and applied to measure their performance on the GA(Genetic Algorithm)features subset and with full features set.The NSL-KDD dataset is used as a benchmark.The empirical results are analyzed,addressed and compared among different activation functions of the ELM.The results show that the radial basis and sine functions perform better on GA feature set than the full feature set while the performance of the sigmoid function is almost equal on both features sets.So,the proposal of GA based feature selection reduced 21 features out of 41 that brought up to 98%accuracy and enhanced overall efficiency of extreme learning machine in intrusion detection.
文摘Metallic alloys for a given application are usually designed to achieve the desired properties by devising experimentsbased on experience, thermodynamic and kinetic principles, and various modeling and simulation exercises.However, the influence of process parameters and material properties is often non-linear and non-colligative. Inrecent years, machine learning (ML) has emerged as a promising tool to dealwith the complex interrelation betweencomposition, properties, and process parameters to facilitate accelerated discovery and development of new alloysand functionalities. In this study, we adopt an ML-based approach, coupled with genetic algorithm (GA) principles,to design novel copper alloys for achieving seemingly contradictory targets of high strength and high electricalconductivity. Initially, we establish a correlation between the alloy composition (binary to multi-component) andthe target properties, namely, electrical conductivity and mechanical strength. Catboost, an ML model coupledwith GA, was used for this task. The accuracy of the model was above 93.5%. Next, for obtaining the optimizedcompositions the outputs fromthe initial model were refined by combining the concepts of data augmentation andPareto front. Finally, the ultimate objective of predicting the target composition that would deliver the desired rangeof properties was achieved by developing an advancedMLmodel through data segregation and data augmentation.To examine the reliability of this model, results were rigorously compared and verified using several independentdata reported in the literature. This comparison substantiates that the results predicted by our model regarding thevariation of conductivity and evolution ofmicrostructure and mechanical properties with composition are in goodagreement with the reports published in the literature.
文摘Extreme Learning Machine(ELM)is popular in batch learning,sequential learning,and progressive learning,due to its speed,easy integration,and generalization ability.While,Traditional ELM cannot train massive data rapidly and efficiently due to its memory residence,high time and space complexity.In ELM,the hidden layer typically necessitates a huge number of nodes.Furthermore,there is no certainty that the arrangement of weights and biases within the hidden layer is optimal.To solve this problem,the traditional ELM has been hybridized with swarm intelligence optimization techniques.This paper displays five proposed hybrid Algorithms“Salp Swarm Algorithm(SSA-ELM),Grasshopper Algorithm(GOA-ELM),Grey Wolf Algorithm(GWO-ELM),Whale optimizationAlgorithm(WOA-ELM)andMoth Flame Optimization(MFO-ELM)”.These five optimizers are hybridized with standard ELM methodology for resolving the tumor type classification using gene expression data.The proposed models applied to the predication of electricity loading data,that describes the energy use of a single residence over a fouryear period.In the hidden layer,Swarm algorithms are used to pick a smaller number of nodes to speed up the execution of ELM.The best weights and preferences were calculated by these algorithms for the hidden layer.Experimental results demonstrated that the proposed MFO-ELM achieved 98.13%accuracy and this is the highest model in accuracy in tumor type classification gene expression data.While in predication,the proposed GOA-ELM achieved 0.397which is least RMSE compared to the other models.
基金This work was supported by National Basic Research Programof China under Grant2002cb312200 01 3National Nature ScienceFoundation of China under Grant60174038.
文摘Support vector machines （SVMs） have been introduced as effective methods for solving classification problems. However, due to some limitations in practical applications, their generalization performance is sometimes far from the expected level. Therefore, it is meaningful to study SVM ensemble learning. In this paper, a novel genetic algorithm based ensemble learning method, namely Direct Genetic Ensemble （DGE）, is proposed. DGE adopts the predictive accuracy of ensemble as the fitness function and searches a good ensemble from the ensemble space. In essence, DGE is also a selective ensemble learning method because the base classifiers of the ensemble are selected according to the solution of genetic algorithm. In comparison with other ensemble learning methods, DGE works on a higher level and is more direct. Different strategies of constructing diverse base classifiers can be utilized in DGE. Experimental results show that SVM ensembles constructed by DGE can achieve better performance than single SVMs, hagged and boosted SVM ensembles. In addition, some valuable conclusions are obtained.
基金The authors extend their appreciation to the Deputyship for Research&Innovation,“Ministry of Education”in Saudi Arabia for funding this research work through the project number (IFKSUDR_D127).
文摘Breast cancer diagnosis through mammography is a pivotal application within medical image-based diagnostics,integral for early detection and effective treatment.While deep learning has significantly advanced the analysis of mammographic images,challenges such as low contrast,image noise,and the high dimensionality of features often degrade model performance.Addressing these challenges,our study introduces a novel method integrating Genetic Algorithms(GA)with pre-trained Convolutional Neural Network(CNN)models to enhance feature selection and classification accuracy.Our approach involves a systematic process:first,we employ widely-used CNN architectures(VGG16,VGG19,MobileNet,and DenseNet)to extract a broad range of features from the Medical Image Analysis Society(MIAS)mammography dataset.Subsequently,a GA optimizes these features by selecting the most relevant and least redundant,aiming to overcome the typical pitfalls of high dimensionality.The selected features are then utilized to train several classifiers,including Linear and Polynomial Support Vector Machines(SVMs),K-Nearest Neighbors,Decision Trees,and Random Forests,enabling a robust evaluation of the method’s effectiveness across varied learning algorithms.Our extensive experimental evaluation demonstrates that the integration of MobileNet and GA significantly improves classification accuracy,from 83.33%to 89.58%,underscoring the method’s efficacy.By detailing these steps,we highlight the innovation of our approach which not only addresses key issues in breast cancer imaging analysis but also offers a scalable solution potentially applicable to other domains within medical imaging.
文摘Neural networks(NNs),as one of the most robust and efficient machine learning methods,have been commonly used in solving several problems.However,choosing proper hyperparameters(e.g.the numbers of layers and neurons in each layer)has a significant influence on the accuracy of these methods.Therefore,a considerable number of studies have been carried out to optimize the NN hyperpaxameters.In this study,the genetic algorithm is applied to NN to find the optimal hyperpaxameters.Thus,the deep energy method,which contains a deep neural network,is applied first on a Timoshenko beam and a plate with a hole.Subsequently,the numbers of hidden layers,integration points,and neurons in each layer are optimized to reach the highest accuracy to predict the stress distribution through these structures.Thus,applying the proper optimization method on NN leads to significant increase in the NN prediction accuracy after conducting the optimization in various examples.
