Multiple faults are easily confused with single faults.In order to identify multiple faults more accurately,a highly efficient learning method is proposed based on a double parallel two-hidden-layer extreme learning m...Multiple faults are easily confused with single faults.In order to identify multiple faults more accurately,a highly efficient learning method is proposed based on a double parallel two-hidden-layer extreme learning machine,called DPTELM.The DPT-ELM method is a variant of an extreme learning machine(ELM).There are some issues with ELM.First,achieving a high accuracy requires too many hidden nodes;second,the direct connection between the input layer and the output layer is ignored.Accordingly,to deal with the above-mentioned problems,DPT-ELM extends the single-hidden-layer ELM to a two-hidden-layer ELM,which can achieve a desired performance with fewer hidden nodes.In addition,a direct connection is built between the input layer and the output layer.Since the input layer weights and the thresholds of the two hidden layers are determined randomly,this simplifies the improved model and shortens the calculation time.Additionally,to improve the signal to noise ratio(SNR),an adaptive waveform decomposition(AWD)algorithm is used to denoise the vibration signal.Then,the denoised signal is used to extract the eigenvalues by the time-domain and frequency-domain methods.Finally,the eigenvalues are input to the DPT-ELM classifier.In this paper,two groups of rolling bearing data at different speeds,which were collected from a real experimental platform,are used to test the method.Each set of data includes three single fault states,two complex fault states and a healthy state.The experimental results demonstrate that the DPT-ELM method achieves fast learning speed and a high accuracy.Moreover,based on 10-fold cross-validation,it proves to be an effective method to improve the accuracy with fewer hidden nodes.展开更多
Lung cancer is among the most frequent cancers in the world,with over one million deaths per year.Classification is required for lung cancer diagnosis and therapy to be effective,accurate,and reliable.Gene expression ...Lung cancer is among the most frequent cancers in the world,with over one million deaths per year.Classification is required for lung cancer diagnosis and therapy to be effective,accurate,and reliable.Gene expression microarrays have made it possible to find genetic biomarkers for cancer diagnosis and prediction in a high-throughput manner.Machine Learning(ML)has been widely used to diagnose and classify lung cancer where the performance of ML methods is evaluated to identify the appropriate technique.Identifying and selecting the gene expression patterns can help in lung cancer diagnoses and classification.Normally,microarrays include several genes and may cause confusion or false prediction.Therefore,the Arithmetic Optimization Algorithm(AOA)is used to identify the optimal gene subset to reduce the number of selected genes.Which can allow the classifiers to yield the best performance for lung cancer classification.In addition,we proposed a modified version of AOA which can work effectively on the high dimensional dataset.In the modified AOA,the features are ranked by their weights and are used to initialize the AOA population.The exploitation process of AOA is then enhanced by developing a local search algorithm based on two neighborhood strategies.Finally,the efficiency of the proposed methods was evaluated on gene expression datasets related to Lung cancer using stratified 4-fold cross-validation.The method’s efficacy in selecting the optimal gene subset is underscored by its ability to maintain feature proportions between 10%to 25%.Moreover,the approach significantly enhances lung cancer prediction accuracy.For instance,Lung_Harvard1 achieved an accuracy of 97.5%,Lung_Harvard2 and Lung_Michigan datasets both achieved 100%,Lung_Adenocarcinoma obtained an accuracy of 88.2%,and Lung_Ontario achieved an accuracy of 87.5%.In conclusion,the results indicate the potential promise of the proposed modified AOA approach in classifying microarray cancer data.展开更多
Adversarial approaches,which intentionally challenge machine learning models by generating difficult examples,are increasingly being adopted to improve machine learning interatomic potentials(MLIPs).While already prov...Adversarial approaches,which intentionally challenge machine learning models by generating difficult examples,are increasingly being adopted to improve machine learning interatomic potentials(MLIPs).While already providing great practical value,little is known about the actual prediction errors of MLIPs on adversarial structures and whether these errors can be controlled.We propose the Calibrated Adversarial Geometry Optimization(CAGO)algorithm to discover adversarial structures with userassigned errors.Through uncertainty calibration,the estimated uncertainty of MLIPs is unified with real errors.By performing geometry optimization for calibrated uncertainty,we reach adversarial structures with the user-assigned target MLIP prediction error.Integrating with active learning pipelines,we benchmark CAGO,demonstrating stable MLIPs that systematically converge structural,dynamical,and thermodynamical properties for liquid water and water adsorption in a metal-organic framework within only hundreds of training structures,where previously many thousands were typically required.展开更多
This paper proposes a novel nondestructive diagnostic method for flip chips based on an improved semi-supervised deep extreme learning machine(ISDELM)and vibration signals.First,an ultrasonic transducer is used to gen...This paper proposes a novel nondestructive diagnostic method for flip chips based on an improved semi-supervised deep extreme learning machine(ISDELM)and vibration signals.First,an ultrasonic transducer is used to generate and focus ultrasounds on the surface of the flip chip to excite it,and a laser scanning vibrometer is applied to acquire the chip’s vibration signals.Then,an extreme learning machine-autoencoder(ELM-AE)structure is adopted to extract features from the original vibration signals layer by layer.Finally,the study proposes integrating the ELM with sparsity neighboring reconstruction to diagnose defects based on unlabeled and labeled data.The ISDELM algorithm is applied to experimental vibration data of flip chips and compared with several other algorithms,such as semi-supervised ELM(SS-ELM),deep ELM,stacked autoencoder,convolutional neural network,and ordinary SDELM.The results show that the proposed method is superior to the several currently available algorithms in terms of accuracy and stability.展开更多
基金supported by National Natural Science Foundation of China(51675035/51375037)
文摘Multiple faults are easily confused with single faults.In order to identify multiple faults more accurately,a highly efficient learning method is proposed based on a double parallel two-hidden-layer extreme learning machine,called DPTELM.The DPT-ELM method is a variant of an extreme learning machine(ELM).There are some issues with ELM.First,achieving a high accuracy requires too many hidden nodes;second,the direct connection between the input layer and the output layer is ignored.Accordingly,to deal with the above-mentioned problems,DPT-ELM extends the single-hidden-layer ELM to a two-hidden-layer ELM,which can achieve a desired performance with fewer hidden nodes.In addition,a direct connection is built between the input layer and the output layer.Since the input layer weights and the thresholds of the two hidden layers are determined randomly,this simplifies the improved model and shortens the calculation time.Additionally,to improve the signal to noise ratio(SNR),an adaptive waveform decomposition(AWD)algorithm is used to denoise the vibration signal.Then,the denoised signal is used to extract the eigenvalues by the time-domain and frequency-domain methods.Finally,the eigenvalues are input to the DPT-ELM classifier.In this paper,two groups of rolling bearing data at different speeds,which were collected from a real experimental platform,are used to test the method.Each set of data includes three single fault states,two complex fault states and a healthy state.The experimental results demonstrate that the DPT-ELM method achieves fast learning speed and a high accuracy.Moreover,based on 10-fold cross-validation,it proves to be an effective method to improve the accuracy with fewer hidden nodes.
基金supported by the Deanship of Scientific Research,at Imam Abdulrahman Bin Faisal University.Grant Number:2019-416-ASCS.
文摘Lung cancer is among the most frequent cancers in the world,with over one million deaths per year.Classification is required for lung cancer diagnosis and therapy to be effective,accurate,and reliable.Gene expression microarrays have made it possible to find genetic biomarkers for cancer diagnosis and prediction in a high-throughput manner.Machine Learning(ML)has been widely used to diagnose and classify lung cancer where the performance of ML methods is evaluated to identify the appropriate technique.Identifying and selecting the gene expression patterns can help in lung cancer diagnoses and classification.Normally,microarrays include several genes and may cause confusion or false prediction.Therefore,the Arithmetic Optimization Algorithm(AOA)is used to identify the optimal gene subset to reduce the number of selected genes.Which can allow the classifiers to yield the best performance for lung cancer classification.In addition,we proposed a modified version of AOA which can work effectively on the high dimensional dataset.In the modified AOA,the features are ranked by their weights and are used to initialize the AOA population.The exploitation process of AOA is then enhanced by developing a local search algorithm based on two neighborhood strategies.Finally,the efficiency of the proposed methods was evaluated on gene expression datasets related to Lung cancer using stratified 4-fold cross-validation.The method’s efficacy in selecting the optimal gene subset is underscored by its ability to maintain feature proportions between 10%to 25%.Moreover,the approach significantly enhances lung cancer prediction accuracy.For instance,Lung_Harvard1 achieved an accuracy of 97.5%,Lung_Harvard2 and Lung_Michigan datasets both achieved 100%,Lung_Adenocarcinoma obtained an accuracy of 88.2%,and Lung_Ontario achieved an accuracy of 87.5%.In conclusion,the results indicate the potential promise of the proposed modified AOA approach in classifying microarray cancer data.
基金supported by the Research Council of Norway through the Centre of Excellence Hylleraas Centre for Quantum Molecular Sciences(Grant 262695)the Young Researcher Talent grants 344993 and 354100+2 种基金We acknowledge the EuroHPC Joint Undertaking for awarding this project access to the EuroHPC supercomputer LUMI,hosted by CSC(Finland)and the LUMI consortium through a EuroHPC Regular Access call(Grants EHPC-REG-2023R02-088,EHPC-REG-2023R03-146)Support was also received from the Centre for Advanced Study in Oslo,Norway,which funded and hosted the SLB Young CAS Fellow research project during the academic year of 23/24 and 24/25Part of the simulations were performed on resources provided by Sigma2—the Norwegian National Infrastructure for High-Performance Computing and Data Storage(grant numbers NN4654K and NS4654K).
文摘Adversarial approaches,which intentionally challenge machine learning models by generating difficult examples,are increasingly being adopted to improve machine learning interatomic potentials(MLIPs).While already providing great practical value,little is known about the actual prediction errors of MLIPs on adversarial structures and whether these errors can be controlled.We propose the Calibrated Adversarial Geometry Optimization(CAGO)algorithm to discover adversarial structures with userassigned errors.Through uncertainty calibration,the estimated uncertainty of MLIPs is unified with real errors.By performing geometry optimization for calibrated uncertainty,we reach adversarial structures with the user-assigned target MLIP prediction error.Integrating with active learning pipelines,we benchmark CAGO,demonstrating stable MLIPs that systematically converge structural,dynamical,and thermodynamical properties for liquid water and water adsorption in a metal-organic framework within only hundreds of training structures,where previously many thousands were typically required.
基金supported by the fellowship of China Postdoctoral Science Foundation(Grant No.2021T140279)the National Natural Science Foundation of China(Grant Nos.51705203,51775243 and 11902124)“111”Project(Grant No.B18027)。
文摘This paper proposes a novel nondestructive diagnostic method for flip chips based on an improved semi-supervised deep extreme learning machine(ISDELM)and vibration signals.First,an ultrasonic transducer is used to generate and focus ultrasounds on the surface of the flip chip to excite it,and a laser scanning vibrometer is applied to acquire the chip’s vibration signals.Then,an extreme learning machine-autoencoder(ELM-AE)structure is adopted to extract features from the original vibration signals layer by layer.Finally,the study proposes integrating the ELM with sparsity neighboring reconstruction to diagnose defects based on unlabeled and labeled data.The ISDELM algorithm is applied to experimental vibration data of flip chips and compared with several other algorithms,such as semi-supervised ELM(SS-ELM),deep ELM,stacked autoencoder,convolutional neural network,and ordinary SDELM.The results show that the proposed method is superior to the several currently available algorithms in terms of accuracy and stability.
