Accurate prediction of flood events is important for flood control and risk management.Machine learning techniques contributed greatly to advances in flood predictions,and existing studies mainly focused on predicting...Accurate prediction of flood events is important for flood control and risk management.Machine learning techniques contributed greatly to advances in flood predictions,and existing studies mainly focused on predicting flood resource variables using single or hybrid machine learning techniques.However,class-based flood predictions have rarely been investigated,which can aid in quickly diagnosing comprehensive flood characteristics and proposing targeted management strategies.This study proposed a prediction approach of flood regime metrics and event classes coupling machine learning algorithms with clustering-deduced membership degrees.Five algorithms were adopted for this exploration.Results showed that the class membership degrees accurately determined event classes with class hit rates up to 100%,compared with the four classes clustered from nine regime metrics.The nonlinear algorithms(Multiple Linear Regression,Random Forest,and least squares-Support Vector Machine)outperformed the linear techniques(Multiple Linear Regression and Stepwise Regression)in predicting flood regime metrics.The proposed approach well predicted flood event classes with average class hit rates of 66.0%-85.4%and 47.2%-76.0%in calibration and validation periods,respectively,particularly for the slow and late flood events.The predictive capability of the proposed prediction approach for flood regime metrics and classes was considerably stronger than that of hydrological modeling approach.展开更多
BACKGROUND Esophageal squamous cell carcinoma is a major histological subtype of esophageal cancer.Many molecular genetic changes are associated with its occurrence.Raman spectroscopy has become a new method for the e...BACKGROUND Esophageal squamous cell carcinoma is a major histological subtype of esophageal cancer.Many molecular genetic changes are associated with its occurrence.Raman spectroscopy has become a new method for the early diagnosis of tumors because it can reflect the structures of substances and their changes at the molecular level.AIM To detect alterations in Raman spectral information across different stages of esophageal neoplasia.METHODS Different grades of esophageal lesions were collected,and a total of 360 groups of Raman spectrum data were collected.A 1D-transformer network model was proposed to handle the task of classifying the spectral data of esophageal squamous cell carcinoma.In addition,a deep learning model was applied to visualize the Raman spectral data and interpret their molecular characteristics.RESULTS A comparison among Raman spectral data with different pathological grades and a visual analysis revealed that the Raman peaks with significant differences were concentrated mainly at 1095 cm^(-1)(DNA,symmetric PO,and stretching vibration),1132 cm^(-1)(cytochrome c),1171 cm^(-1)(acetoacetate),1216 cm^(-1)(amide III),and 1315 cm^(-1)(glycerol).A comparison among the training results of different models revealed that the 1Dtransformer network performed best.A 93.30%accuracy value,a 96.65%specificity value,a 93.30%sensitivity value,and a 93.17%F1 score were achieved.CONCLUSION Raman spectroscopy revealed significantly different waveforms for the different stages of esophageal neoplasia.The combination of Raman spectroscopy and deep learning methods could significantly improve the accuracy of classification.展开更多
Neuromorphic computing extends beyond sequential processing modalities and outperforms traditional von Neumann architectures in implementing more complicated tasks,e.g.,pattern processing,image recognition,and decisio...Neuromorphic computing extends beyond sequential processing modalities and outperforms traditional von Neumann architectures in implementing more complicated tasks,e.g.,pattern processing,image recognition,and decision making.It features parallel interconnected neural networks,high fault tolerance,robustness,autonomous learning capability,and ultralow energy dissipation.The algorithms of artificial neural network(ANN)have also been widely used because of their facile self-organization and self-learning capabilities,which mimic those of the human brain.To some extent,ANN reflects several basic functions of the human brain and can be efficiently integrated into neuromorphic devices to perform neuromorphic computations.This review highlights recent advances in neuromorphic devices assisted by machine learning algorithms.First,the basic structure of simple neuron models inspired by biological neurons and the information processing in simple neural networks are particularly discussed.Second,the fabrication and research progress of neuromorphic devices are presented regarding to materials and structures.Furthermore,the fabrication of neuromorphic devices,including stand-alone neuromorphic devices,neuromorphic device arrays,and integrated neuromorphic systems,is discussed and demonstrated with reference to some respective studies.The applications of neuromorphic devices assisted by machine learning algorithms in different fields are categorized and investigated.Finally,perspectives,suggestions,and potential solutions to the current challenges of neuromorphic devices are provided.展开更多
Based on the Google Earth Engine cloud computing data platform,this study employed three algorithms including Support Vector Machine,Random Forest,and Classification and Regression Tree to classify the current status ...Based on the Google Earth Engine cloud computing data platform,this study employed three algorithms including Support Vector Machine,Random Forest,and Classification and Regression Tree to classify the current status of land covers in Hung Yen province of Vietnam using Landsat 8 OLI satellite images,a free data source with reasonable spatial and temporal resolution.The results of the study show that all three algorithms presented good classification for five basic types of land cover including Rice land,Water bodies,Perennial vegetation,Annual vegetation,Built-up areas as their overall accuracy and Kappa coefficient were greater than 80%and 0.8,respectively.Among the three algorithms,SVM achieved the highest accuracy as its overall accuracy was 86%and the Kappa coefficient was 0.88.Land cover classification based on the SVM algorithm shows that Built-up areas cover the largest area with nearly 31,495 ha,accounting for more than 33.8%of the total natural area,followed by Rice land and Perennial vegetation which cover an area of over 30,767 ha(33%)and 15,637 ha(16.8%),respectively.Water bodies and Annual vegetation cover the smallest areas with 8,820(9.5%)ha and 6,302 ha(6.8%),respectively.The results of this study can be used for land use management and planning as well as other natural resource and environmental management purposes in the province.展开更多
The distillation process is an important chemical process,and the application of data-driven modelling approach has the potential to reduce model complexity compared to mechanistic modelling,thus improving the efficie...The distillation process is an important chemical process,and the application of data-driven modelling approach has the potential to reduce model complexity compared to mechanistic modelling,thus improving the efficiency of process optimization or monitoring studies.However,the distillation process is highly nonlinear and has multiple uncertainty perturbation intervals,which brings challenges to accurate data-driven modelling of distillation processes.This paper proposes a systematic data-driven modelling framework to solve these problems.Firstly,data segment variance was introduced into the K-means algorithm to form K-means data interval(KMDI)clustering in order to cluster the data into perturbed and steady state intervals for steady-state data extraction.Secondly,maximal information coefficient(MIC)was employed to calculate the nonlinear correlation between variables for removing redundant features.Finally,extreme gradient boosting(XGBoost)was integrated as the basic learner into adaptive boosting(AdaBoost)with the error threshold(ET)set to improve weights update strategy to construct the new integrated learning algorithm,XGBoost-AdaBoost-ET.The superiority of the proposed framework is verified by applying this data-driven modelling framework to a real industrial process of propylene distillation.展开更多
BACKGROUND Difficulty of colonoscopy insertion(DCI)significantly affects colonoscopy effectiveness and serves as a key quality indicator.Predicting and evaluating DCI risk preoperatively is crucial for optimizing intr...BACKGROUND Difficulty of colonoscopy insertion(DCI)significantly affects colonoscopy effectiveness and serves as a key quality indicator.Predicting and evaluating DCI risk preoperatively is crucial for optimizing intraoperative strategies.AIM To evaluate the predictive performance of machine learning(ML)algorithms for DCI by comparing three modeling approaches,identify factors influencing DCI,and develop a preoperative prediction model using ML algorithms to enhance colonoscopy quality and efficiency.METHODS This cross-sectional study enrolled 712 patients who underwent colonoscopy at a tertiary hospital between June 2020 and May 2021.Demographic data,past medical history,medication use,and psychological status were collected.The endoscopist assessed DCI using the visual analogue scale.After univariate screening,predictive models were developed using multivariable logistic regression,least absolute shrinkage and selection operator(LASSO)regression,and random forest(RF)algorithms.Model performance was evaluated based on discrimination,calibration,and decision curve analysis(DCA),and results were visualized using nomograms.RESULTS A total of 712 patients(53.8%male;mean age 54.5 years±12.9 years)were included.Logistic regression analysis identified constipation[odds ratio(OR)=2.254,95%confidence interval(CI):1.289-3.931],abdominal circumference(AC)(77.5–91.9 cm,OR=1.895,95%CI:1.065-3.350;AC≥92 cm,OR=1.271,95%CI:0.730-2.188),and anxiety(OR=1.071,95%CI:1.044-1.100)as predictive factors for DCI,validated by LASSO and RF methods.Model performance revealed training/validation sensitivities of 0.826/0.925,0.924/0.868,and 1.000/0.981;specificities of 0.602/0.511,0.510/0.562,and 0.977/0.526;and corresponding area under the receiver operating characteristic curves(AUCs)of 0.780(0.737-0.823)/0.726(0.654-0.799),0.754(0.710-0.798)/0.723(0.656-0.791),and 1.000(1.000-1.000)/0.754(0.688-0.820),respectively.DCA indicated optimal net benefit within probability thresholds of 0-0.9 and 0.05-0.37.The RF model demonstrated superior diagnostic accuracy,reflected by perfect training sensitivity(1.000)and highest validation AUC(0.754),outperforming other methods in clinical applicability.CONCLUSION The RF-based model exhibited superior predictive accuracy for DCI compared to multivariable logistic and LASSO regression models.This approach supports individualized preoperative optimization,enhancing colonoscopy quality through targeted risk stratification.展开更多
DNA microarrays, a cornerstone in biomedicine, measure gene expression across thousands to tens of thousands of genes. Identifying the genes vital for accurate cancer classification is a key challenge. Here, we presen...DNA microarrays, a cornerstone in biomedicine, measure gene expression across thousands to tens of thousands of genes. Identifying the genes vital for accurate cancer classification is a key challenge. Here, we present Fs-LSA (F-score based Learning Search Algorithm), a novel gene selection algorithm designed to enhance the precision and efficiency of target gene identification from microarray data for cancer classification. This algorithm is divided into two phases: the first leverages F-score values to prioritize and select feature genes with the most significant differential expression;the second phase introduces our Learning Search Algorithm (LSA), which harnesses swarm intelligence to identify the optimal subset among the remaining genes. Inspired by human social learning, LSA integrates historical data and collective intelligence for a thorough search, with a dynamic control mechanism that balances exploration and refinement, thereby enhancing the gene selection process. We conducted a rigorous validation of Fs-LSA’s performance using eight publicly available cancer microarray expression datasets. Fs-LSA achieved accuracy, precision, sensitivity, and F1-score values of 0.9932, 0.9923, 0.9962, and 0.994, respectively. Comparative analyses with state-of-the-art algorithms revealed Fs-LSA’s superior performance in terms of simplicity and efficiency. Additionally, we validated the algorithm’s efficacy independently using glioblastoma data from GEO and TCGA databases. It was significantly superior to those of the comparison algorithms. Importantly, the driver genes identified by Fs-LSA were instrumental in developing a predictive model as an independent prognostic indicator for glioblastoma, underscoring Fs-LSA’s transformative potential in genomics and personalized medicine.展开更多
A learning algorithm based on a hard limiter for feedforward neural networks (NN) is presented,and is applied in solving classification problems on separable convex sets and disjoint sets.It has been proved that the a...A learning algorithm based on a hard limiter for feedforward neural networks (NN) is presented,and is applied in solving classification problems on separable convex sets and disjoint sets.It has been proved that the algorithm has stronger classification ability than that of the back propagation (BP) algorithm for the feedforward NN using sigmoid function by simulation.What is more,the models can be implemented with lower cost hardware than that of the BP NN.LEARNIN展开更多
To overcome the limitation that complex data types with noun attributes cannot be processed by rank learning algorithms, a new rank learning algorithm is designed. In the learning algorithm based on the decision tree,...To overcome the limitation that complex data types with noun attributes cannot be processed by rank learning algorithms, a new rank learning algorithm is designed. In the learning algorithm based on the decision tree, the splitting rule of the decision tree is revised with a new definition of rank impurity. A new rank learning algorithm, which can be intuitively explained, is obtained and its theoretical basis is provided. The experimental results show that in the aspect of average rank loss, the ranking tree algorithm outperforms perception ranking and ordinal regression algorithms and it also has a faster convergence speed. The rank learning algorithm based on the decision tree is able to process categorical data and select relative features.展开更多
The current study aimed at evaluating the capabilities of seven advanced machine learning techniques(MLTs),including,Support Vector Machine(SVM),Random Forest(RF),Multivariate Adaptive Regression Spline(MARS),Artifici...The current study aimed at evaluating the capabilities of seven advanced machine learning techniques(MLTs),including,Support Vector Machine(SVM),Random Forest(RF),Multivariate Adaptive Regression Spline(MARS),Artificial Neural Network(ANN),Quadratic Discriminant Analysis(QDA),Linear Discriminant Analysis(LDA),and Naive Bayes(NB),for landslide susceptibility modeling and comparison of their performances.Coupling machine learning algorithms with spatial data types for landslide susceptibility mapping is a vitally important issue.This study was carried out using GIS and R open source software at Abha Basin,Asir Region,Saudi Arabia.First,a total of 243 landslide locations were identified at Abha Basin to prepare the landslide inventory map using different data sources.All the landslide areas were randomly separated into two groups with a ratio of 70%for training and 30%for validating purposes.Twelve landslide-variables were generated for landslide susceptibility modeling,which include altitude,lithology,distance to faults,normalized difference vegetation index(NDVI),landuse/landcover(LULC),distance to roads,slope angle,distance to streams,profile curvature,plan curvature,slope length(LS),and slope-aspect.The area under curve(AUC-ROC)approach has been applied to evaluate,validate,and compare the MLTs performance.The results indicated that AUC values for seven MLTs range from 89.0%for QDA to 95.1%for RF.Our findings showed that the RF(AUC=95.1%)and LDA(AUC=941.7%)have produced the best performances in comparison to other MLTs.The outcome of this study and the landslide susceptibility maps would be useful for environmental protection.展开更多
Due to the combined influences such as ore-forming temperature,fluid and metal sources,sphalerite tends to incorporate diverse contents of trace elements during the formation of different types of Lead-zinc(Pb-Zn)depo...Due to the combined influences such as ore-forming temperature,fluid and metal sources,sphalerite tends to incorporate diverse contents of trace elements during the formation of different types of Lead-zinc(Pb-Zn)deposits.Therefore,trace elements in sphalerite have long been utilized to distinguish Pb-Zn deposit types.However,previous discriminant diagrams usually contain two or three dimensions,which are limited to revealing the complicated interrelations between trace elements of sphalerite and the types of Pb-Zn deposits.In this study,we aim to prove that the sphalerite trace elements can be used to classify the Pb-Zn deposit types and extract key factors from sphalerite trace elements that can dis-criminate Pb-Zn deposit types using machine learning algorithms.A dataset of nearly 3600 sphalerite spot analyses from 95 Pb-Zn deposits worldwide determined by LA-ICP-MS was compiled from peer-reviewed publications,containing 12 elements(Mn,Fe,Co,Cu,Ga,Ge,Ag,Cd,In,Sn,Sb,and Pb)from 5 types,including Sedimentary Exhalative(SEDEX),Mississippi Valley Type(MVT),Volcanic Massive Sulfide(VMS),skarn,and epithermal deposits.Random Forests(RF)is applied to the data processing and the results show that trace elements of sphalerite can successfully discriminate different types of Pb-Zn deposits except for VMS deposits,most of which are falsely distinguished as skarn and epithermal types.To further discriminate VMS deposits,future studies could focus on enlarging the capacity of VMS deposits in datasets and applying other geological factors along with sphalerite trace elements when con-structing the classification model.RF’s feature importance and permutation feature importance were adopted to evaluate the element significance for classification.Besides,a visualized tool,t-distributed stochastic neighbor embedding(t-SNE),was used to verify the results of both classification and evalua-tion.The results presented here show that Mn,Co,and Ge display significant impacts on classification of Pb-Zn deposits and In,Ga,Sn,Cd,and Fe also have relatively important effects compared to the rest ele-ments,confirming that Pb-Zn deposits discrimination is mainly controlled by multi-elements in spha-lerite.Our study hence shows that machine learning algorithm can provide new insights into conventional geochemical analyses,inspiring future research on constructing classification models of mineral deposits using mineral geochemistry data.展开更多
The risk of rockbursts is one of the main threats in hard coal mines. Compared to other underground mines, the number of factors contributing to the rockburst at underground coal mines is much greater.Factors such as ...The risk of rockbursts is one of the main threats in hard coal mines. Compared to other underground mines, the number of factors contributing to the rockburst at underground coal mines is much greater.Factors such as the coal seam tendency to rockbursts, the thickness of the coal seam, and the stress level in the seam have to be considered, but also the entire coal seam-surrounding rock system has to be evaluated when trying to predict the rockbursts. However, in hard coal mines, there are stroke or stress-stroke rockbursts in which the fracture of a thick layer of sandstone plays an essential role in predicting rockbursts. The occurrence of rockbursts in coal mines is complex, and their prediction is even more difficult than in other mines. In recent years, the interest in machine learning algorithms for solving complex nonlinear problems has increased, which also applies to geosciences. This study attempts to use machine learning algorithms, i.e. neural network, decision tree, random forest, gradient boosting, and extreme gradient boosting(XGB), to assess the rockburst hazard of an active hard coal mine in the Upper Silesian Coal Basin. The rock mass bursting tendency index WTGthat describes the tendency of the seam-surrounding rock system to rockbursts and the anomaly of the vertical stress component were applied for this purpose. Especially, the decision tree and neural network models were proved to be effective in correctly distinguishing rockbursts from tremors, after which the excavation was not damaged. On average, these models correctly classified about 80% of the rockbursts in the testing datasets.展开更多
In this paper, we explore a novel ensemble method for spectral clustering. In contrast to the traditional clustering ensemble methods that combine all the obtained clustering results, we propose the adaptive spectral ...In this paper, we explore a novel ensemble method for spectral clustering. In contrast to the traditional clustering ensemble methods that combine all the obtained clustering results, we propose the adaptive spectral clustering ensemble method to achieve a better clustering solution. This method can adaptively assess the number of the component members, which is not owned by many other algorithms. The component clusterings of the ensemble system are generated by spectral clustering （SC） which bears some good characteristics to engender the diverse committees. The selection process works by evaluating the generated component spectral clustering through resampling technique and population-based incremental learning algorithm （PBIL）. Experimental results on UCI datasets demonstrate that the proposed algorithm can achieve better results compared with traditional clustering ensemble methods, especially when the number of component clusterings is large.展开更多
This investigation assessed the efficacy of 10 widely used machine learning algorithms(MLA)comprising the least absolute shrinkage and selection operator(LASSO),generalized linear model(GLM),stepwise generalized linea...This investigation assessed the efficacy of 10 widely used machine learning algorithms(MLA)comprising the least absolute shrinkage and selection operator(LASSO),generalized linear model(GLM),stepwise generalized linear model(SGLM),elastic net(ENET),partial least square(PLS),ridge regression,support vector machine(SVM),classification and regression trees(CART),bagged CART,and random forest(RF)for gully erosion susceptibility mapping(GESM)in Iran.The location of 462 previously existing gully erosion sites were mapped through widespread field investigations,of which 70%(323)and 30%(139)of observations were arbitrarily divided for algorithm calibration and validation.Twelve controlling factors for gully erosion,namely,soil texture,annual mean rainfall,digital elevation model(DEM),drainage density,slope,lithology,topographic wetness index(TWI),distance from rivers,aspect,distance from roads,plan curvature,and profile curvature were ranked in terms of their importance using each MLA.The MLA were compared using a training dataset for gully erosion and statistical measures such as RMSE(root mean square error),MAE(mean absolute error),and R-squared.Based on the comparisons among MLA,the RF algorithm exhibited the minimum RMSE and MAE and the maximum value of R-squared,and was therefore selected as the best model.The variable importance evaluation using the RF model revealed that distance from rivers had the highest significance in influencing the occurrence of gully erosion whereas plan curvature had the least importance.According to the GESM generated using RF,most of the study area is predicted to have a low(53.72%)or moderate(29.65%)susceptibility to gully erosion,whereas only a small area is identified to have a high(12.56%)or very high(4.07%)susceptibility.The outcome generated by RF model is validated using the ROC(Receiver Operating Characteristics)curve approach,which returned an area under the curve(AUC)of 0.985,proving the excellent forecasting ability of the model.The GESM prepared using the RF algorithm can aid decision-makers in targeting remedial actions for minimizing the damage caused by gully erosion.展开更多
The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are...The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are inferred based on the maximum likelihood rule. Secondly, the forgetting factor and learning rate factor are redefined, and their still more general formulations are obtained by analyzing their practical functions. Lastly, the convergence of the proposed algorithm is proved to enable the estimation converge to a local maximum of the data likelihood function according to the stochastic approximation theory. The experiments show that the proposed learning algorithm excels the formers both in converging rate and accuracy.展开更多
Big data analytic techniques associated with machine learning algorithms are playing an increasingly important role in various application fields,including stock market investment.However,few studies have focused on f...Big data analytic techniques associated with machine learning algorithms are playing an increasingly important role in various application fields,including stock market investment.However,few studies have focused on forecasting daily stock market returns,especially when using powerful machine learning techniques,such as deep neural networks(DNNs),to perform the analyses.DNNs employ various deep learning algorithms based on the combination of network structure,activation function,and model parameters,with their performance depending on the format of the data representation.This paper presents a comprehensive big data analytics process to predict the daily return direction of the SPDR S&P 500 ETF(ticker symbol:SPY)based on 60 financial and economic features.DNNs and traditional artificial neural networks(ANNs)are then deployed over the entire preprocessed but untransformed dataset,along with two datasets transformed via principal component analysis(PCA),to predict the daily direction of future stock market index returns.While controlling for overfitting,a pattern for the classification accuracy of the DNNs is detected and demonstrated as the number of the hidden layers increases gradually from 12 to 1000.Moreover,a set of hypothesis testing procedures are implemented on the classification,and the simulation results show that the DNNs using two PCA-represented datasets give significantly higher classification accuracy than those using the entire untransformed dataset,as well as several other hybrid machine learning algorithms.In addition,the trading strategies guided by the DNN classification process based on PCA-represented data perform slightly better than the others tested,including in a comparison against two standard benchmarks.展开更多
基金National Key Research and Development Program of China,No.2023YFC3006704National Natural Science Foundation of China,No.42171047CAS-CSIRO Partnership Joint Project of 2024,No.177GJHZ2023097MI。
文摘Accurate prediction of flood events is important for flood control and risk management.Machine learning techniques contributed greatly to advances in flood predictions,and existing studies mainly focused on predicting flood resource variables using single or hybrid machine learning techniques.However,class-based flood predictions have rarely been investigated,which can aid in quickly diagnosing comprehensive flood characteristics and proposing targeted management strategies.This study proposed a prediction approach of flood regime metrics and event classes coupling machine learning algorithms with clustering-deduced membership degrees.Five algorithms were adopted for this exploration.Results showed that the class membership degrees accurately determined event classes with class hit rates up to 100%,compared with the four classes clustered from nine regime metrics.The nonlinear algorithms(Multiple Linear Regression,Random Forest,and least squares-Support Vector Machine)outperformed the linear techniques(Multiple Linear Regression and Stepwise Regression)in predicting flood regime metrics.The proposed approach well predicted flood event classes with average class hit rates of 66.0%-85.4%and 47.2%-76.0%in calibration and validation periods,respectively,particularly for the slow and late flood events.The predictive capability of the proposed prediction approach for flood regime metrics and classes was considerably stronger than that of hydrological modeling approach.
基金Supported by Beijing Hospitals Authority Youth Programme,No.QML20200505.
文摘BACKGROUND Esophageal squamous cell carcinoma is a major histological subtype of esophageal cancer.Many molecular genetic changes are associated with its occurrence.Raman spectroscopy has become a new method for the early diagnosis of tumors because it can reflect the structures of substances and their changes at the molecular level.AIM To detect alterations in Raman spectral information across different stages of esophageal neoplasia.METHODS Different grades of esophageal lesions were collected,and a total of 360 groups of Raman spectrum data were collected.A 1D-transformer network model was proposed to handle the task of classifying the spectral data of esophageal squamous cell carcinoma.In addition,a deep learning model was applied to visualize the Raman spectral data and interpret their molecular characteristics.RESULTS A comparison among Raman spectral data with different pathological grades and a visual analysis revealed that the Raman peaks with significant differences were concentrated mainly at 1095 cm^(-1)(DNA,symmetric PO,and stretching vibration),1132 cm^(-1)(cytochrome c),1171 cm^(-1)(acetoacetate),1216 cm^(-1)(amide III),and 1315 cm^(-1)(glycerol).A comparison among the training results of different models revealed that the 1Dtransformer network performed best.A 93.30%accuracy value,a 96.65%specificity value,a 93.30%sensitivity value,and a 93.17%F1 score were achieved.CONCLUSION Raman spectroscopy revealed significantly different waveforms for the different stages of esophageal neoplasia.The combination of Raman spectroscopy and deep learning methods could significantly improve the accuracy of classification.
基金financially supported by the National Natural Science Foundation of China(No.52073031)the National Key Research and Development Program of China(Nos.2023YFB3208102,2021YFB3200304)+4 种基金the China National Postdoctoral Program for Innovative Talents(No.BX2021302)the Beijing Nova Program(Nos.Z191100001119047,Z211100002121148)the Fundamental Research Funds for the Central Universities(No.E0EG6801X2)the‘Hundred Talents Program’of the Chinese Academy of Sciencesthe BrainLink program funded by the MSIT through the NRF of Korea(No.RS-2023-00237308).
文摘Neuromorphic computing extends beyond sequential processing modalities and outperforms traditional von Neumann architectures in implementing more complicated tasks,e.g.,pattern processing,image recognition,and decision making.It features parallel interconnected neural networks,high fault tolerance,robustness,autonomous learning capability,and ultralow energy dissipation.The algorithms of artificial neural network(ANN)have also been widely used because of their facile self-organization and self-learning capabilities,which mimic those of the human brain.To some extent,ANN reflects several basic functions of the human brain and can be efficiently integrated into neuromorphic devices to perform neuromorphic computations.This review highlights recent advances in neuromorphic devices assisted by machine learning algorithms.First,the basic structure of simple neuron models inspired by biological neurons and the information processing in simple neural networks are particularly discussed.Second,the fabrication and research progress of neuromorphic devices are presented regarding to materials and structures.Furthermore,the fabrication of neuromorphic devices,including stand-alone neuromorphic devices,neuromorphic device arrays,and integrated neuromorphic systems,is discussed and demonstrated with reference to some respective studies.The applications of neuromorphic devices assisted by machine learning algorithms in different fields are categorized and investigated.Finally,perspectives,suggestions,and potential solutions to the current challenges of neuromorphic devices are provided.
文摘Based on the Google Earth Engine cloud computing data platform,this study employed three algorithms including Support Vector Machine,Random Forest,and Classification and Regression Tree to classify the current status of land covers in Hung Yen province of Vietnam using Landsat 8 OLI satellite images,a free data source with reasonable spatial and temporal resolution.The results of the study show that all three algorithms presented good classification for five basic types of land cover including Rice land,Water bodies,Perennial vegetation,Annual vegetation,Built-up areas as their overall accuracy and Kappa coefficient were greater than 80%and 0.8,respectively.Among the three algorithms,SVM achieved the highest accuracy as its overall accuracy was 86%and the Kappa coefficient was 0.88.Land cover classification based on the SVM algorithm shows that Built-up areas cover the largest area with nearly 31,495 ha,accounting for more than 33.8%of the total natural area,followed by Rice land and Perennial vegetation which cover an area of over 30,767 ha(33%)and 15,637 ha(16.8%),respectively.Water bodies and Annual vegetation cover the smallest areas with 8,820(9.5%)ha and 6,302 ha(6.8%),respectively.The results of this study can be used for land use management and planning as well as other natural resource and environmental management purposes in the province.
基金supported by the National Key Research and Development Program of China(2023YFB3307801)the National Natural Science Foundation of China(62394343,62373155,62073142)+3 种基金Major Science and Technology Project of Xinjiang(No.2022A01006-4)the Programme of Introducing Talents of Discipline to Universities(the 111 Project)under Grant B17017the Fundamental Research Funds for the Central Universities,Science Foundation of China University of Petroleum,Beijing(No.2462024YJRC011)the Open Research Project of the State Key Laboratory of Industrial Control Technology,China(Grant No.ICT2024B70).
文摘The distillation process is an important chemical process,and the application of data-driven modelling approach has the potential to reduce model complexity compared to mechanistic modelling,thus improving the efficiency of process optimization or monitoring studies.However,the distillation process is highly nonlinear and has multiple uncertainty perturbation intervals,which brings challenges to accurate data-driven modelling of distillation processes.This paper proposes a systematic data-driven modelling framework to solve these problems.Firstly,data segment variance was introduced into the K-means algorithm to form K-means data interval(KMDI)clustering in order to cluster the data into perturbed and steady state intervals for steady-state data extraction.Secondly,maximal information coefficient(MIC)was employed to calculate the nonlinear correlation between variables for removing redundant features.Finally,extreme gradient boosting(XGBoost)was integrated as the basic learner into adaptive boosting(AdaBoost)with the error threshold(ET)set to improve weights update strategy to construct the new integrated learning algorithm,XGBoost-AdaBoost-ET.The superiority of the proposed framework is verified by applying this data-driven modelling framework to a real industrial process of propylene distillation.
基金the Chinese Clinical Trial Registry(No.ChiCTR2000040109)approved by the Hospital Ethics Committee(No.20210130017).
文摘BACKGROUND Difficulty of colonoscopy insertion(DCI)significantly affects colonoscopy effectiveness and serves as a key quality indicator.Predicting and evaluating DCI risk preoperatively is crucial for optimizing intraoperative strategies.AIM To evaluate the predictive performance of machine learning(ML)algorithms for DCI by comparing three modeling approaches,identify factors influencing DCI,and develop a preoperative prediction model using ML algorithms to enhance colonoscopy quality and efficiency.METHODS This cross-sectional study enrolled 712 patients who underwent colonoscopy at a tertiary hospital between June 2020 and May 2021.Demographic data,past medical history,medication use,and psychological status were collected.The endoscopist assessed DCI using the visual analogue scale.After univariate screening,predictive models were developed using multivariable logistic regression,least absolute shrinkage and selection operator(LASSO)regression,and random forest(RF)algorithms.Model performance was evaluated based on discrimination,calibration,and decision curve analysis(DCA),and results were visualized using nomograms.RESULTS A total of 712 patients(53.8%male;mean age 54.5 years±12.9 years)were included.Logistic regression analysis identified constipation[odds ratio(OR)=2.254,95%confidence interval(CI):1.289-3.931],abdominal circumference(AC)(77.5–91.9 cm,OR=1.895,95%CI:1.065-3.350;AC≥92 cm,OR=1.271,95%CI:0.730-2.188),and anxiety(OR=1.071,95%CI:1.044-1.100)as predictive factors for DCI,validated by LASSO and RF methods.Model performance revealed training/validation sensitivities of 0.826/0.925,0.924/0.868,and 1.000/0.981;specificities of 0.602/0.511,0.510/0.562,and 0.977/0.526;and corresponding area under the receiver operating characteristic curves(AUCs)of 0.780(0.737-0.823)/0.726(0.654-0.799),0.754(0.710-0.798)/0.723(0.656-0.791),and 1.000(1.000-1.000)/0.754(0.688-0.820),respectively.DCA indicated optimal net benefit within probability thresholds of 0-0.9 and 0.05-0.37.The RF model demonstrated superior diagnostic accuracy,reflected by perfect training sensitivity(1.000)and highest validation AUC(0.754),outperforming other methods in clinical applicability.CONCLUSION The RF-based model exhibited superior predictive accuracy for DCI compared to multivariable logistic and LASSO regression models.This approach supports individualized preoperative optimization,enhancing colonoscopy quality through targeted risk stratification.
基金supported by the National Natural Science Foundation of China(Grant Number 62341210)Natural Science Foundation of Guangxi Province(Grant Number:2025GXNSFHA069267)Science and Technology Development Plan for Baise City(Grant Number 20233654).
文摘DNA microarrays, a cornerstone in biomedicine, measure gene expression across thousands to tens of thousands of genes. Identifying the genes vital for accurate cancer classification is a key challenge. Here, we present Fs-LSA (F-score based Learning Search Algorithm), a novel gene selection algorithm designed to enhance the precision and efficiency of target gene identification from microarray data for cancer classification. This algorithm is divided into two phases: the first leverages F-score values to prioritize and select feature genes with the most significant differential expression;the second phase introduces our Learning Search Algorithm (LSA), which harnesses swarm intelligence to identify the optimal subset among the remaining genes. Inspired by human social learning, LSA integrates historical data and collective intelligence for a thorough search, with a dynamic control mechanism that balances exploration and refinement, thereby enhancing the gene selection process. We conducted a rigorous validation of Fs-LSA’s performance using eight publicly available cancer microarray expression datasets. Fs-LSA achieved accuracy, precision, sensitivity, and F1-score values of 0.9932, 0.9923, 0.9962, and 0.994, respectively. Comparative analyses with state-of-the-art algorithms revealed Fs-LSA’s superior performance in terms of simplicity and efficiency. Additionally, we validated the algorithm’s efficacy independently using glioblastoma data from GEO and TCGA databases. It was significantly superior to those of the comparison algorithms. Importantly, the driver genes identified by Fs-LSA were instrumental in developing a predictive model as an independent prognostic indicator for glioblastoma, underscoring Fs-LSA’s transformative potential in genomics and personalized medicine.
文摘A learning algorithm based on a hard limiter for feedforward neural networks (NN) is presented,and is applied in solving classification problems on separable convex sets and disjoint sets.It has been proved that the algorithm has stronger classification ability than that of the back propagation (BP) algorithm for the feedforward NN using sigmoid function by simulation.What is more,the models can be implemented with lower cost hardware than that of the BP NN.LEARNIN
基金The Planning Program of Science and Technology of Hunan Province (No05JT1039)
文摘To overcome the limitation that complex data types with noun attributes cannot be processed by rank learning algorithms, a new rank learning algorithm is designed. In the learning algorithm based on the decision tree, the splitting rule of the decision tree is revised with a new definition of rank impurity. A new rank learning algorithm, which can be intuitively explained, is obtained and its theoretical basis is provided. The experimental results show that in the aspect of average rank loss, the ranking tree algorithm outperforms perception ranking and ordinal regression algorithms and it also has a faster convergence speed. The rank learning algorithm based on the decision tree is able to process categorical data and select relative features.
文摘The current study aimed at evaluating the capabilities of seven advanced machine learning techniques(MLTs),including,Support Vector Machine(SVM),Random Forest(RF),Multivariate Adaptive Regression Spline(MARS),Artificial Neural Network(ANN),Quadratic Discriminant Analysis(QDA),Linear Discriminant Analysis(LDA),and Naive Bayes(NB),for landslide susceptibility modeling and comparison of their performances.Coupling machine learning algorithms with spatial data types for landslide susceptibility mapping is a vitally important issue.This study was carried out using GIS and R open source software at Abha Basin,Asir Region,Saudi Arabia.First,a total of 243 landslide locations were identified at Abha Basin to prepare the landslide inventory map using different data sources.All the landslide areas were randomly separated into two groups with a ratio of 70%for training and 30%for validating purposes.Twelve landslide-variables were generated for landslide susceptibility modeling,which include altitude,lithology,distance to faults,normalized difference vegetation index(NDVI),landuse/landcover(LULC),distance to roads,slope angle,distance to streams,profile curvature,plan curvature,slope length(LS),and slope-aspect.The area under curve(AUC-ROC)approach has been applied to evaluate,validate,and compare the MLTs performance.The results indicated that AUC values for seven MLTs range from 89.0%for QDA to 95.1%for RF.Our findings showed that the RF(AUC=95.1%)and LDA(AUC=941.7%)have produced the best performances in comparison to other MLTs.The outcome of this study and the landslide susceptibility maps would be useful for environmental protection.
基金We would like to acknowledge the financial support of the Ministry of Science and Technology of China(Grant No.2021YFC2900300)the National Natural Science Foundation of China(Grant Nos.41772074 and 42172103).
文摘Due to the combined influences such as ore-forming temperature,fluid and metal sources,sphalerite tends to incorporate diverse contents of trace elements during the formation of different types of Lead-zinc(Pb-Zn)deposits.Therefore,trace elements in sphalerite have long been utilized to distinguish Pb-Zn deposit types.However,previous discriminant diagrams usually contain two or three dimensions,which are limited to revealing the complicated interrelations between trace elements of sphalerite and the types of Pb-Zn deposits.In this study,we aim to prove that the sphalerite trace elements can be used to classify the Pb-Zn deposit types and extract key factors from sphalerite trace elements that can dis-criminate Pb-Zn deposit types using machine learning algorithms.A dataset of nearly 3600 sphalerite spot analyses from 95 Pb-Zn deposits worldwide determined by LA-ICP-MS was compiled from peer-reviewed publications,containing 12 elements(Mn,Fe,Co,Cu,Ga,Ge,Ag,Cd,In,Sn,Sb,and Pb)from 5 types,including Sedimentary Exhalative(SEDEX),Mississippi Valley Type(MVT),Volcanic Massive Sulfide(VMS),skarn,and epithermal deposits.Random Forests(RF)is applied to the data processing and the results show that trace elements of sphalerite can successfully discriminate different types of Pb-Zn deposits except for VMS deposits,most of which are falsely distinguished as skarn and epithermal types.To further discriminate VMS deposits,future studies could focus on enlarging the capacity of VMS deposits in datasets and applying other geological factors along with sphalerite trace elements when con-structing the classification model.RF’s feature importance and permutation feature importance were adopted to evaluate the element significance for classification.Besides,a visualized tool,t-distributed stochastic neighbor embedding(t-SNE),was used to verify the results of both classification and evalua-tion.The results presented here show that Mn,Co,and Ge display significant impacts on classification of Pb-Zn deposits and In,Ga,Sn,Cd,and Fe also have relatively important effects compared to the rest ele-ments,confirming that Pb-Zn deposits discrimination is mainly controlled by multi-elements in spha-lerite.Our study hence shows that machine learning algorithm can provide new insights into conventional geochemical analyses,inspiring future research on constructing classification models of mineral deposits using mineral geochemistry data.
基金supported by the Ministry of Science and Higher Education, Republic of Poland (Statutory Activity of the Central Mining Institute, Grant No. 11133010)
文摘The risk of rockbursts is one of the main threats in hard coal mines. Compared to other underground mines, the number of factors contributing to the rockburst at underground coal mines is much greater.Factors such as the coal seam tendency to rockbursts, the thickness of the coal seam, and the stress level in the seam have to be considered, but also the entire coal seam-surrounding rock system has to be evaluated when trying to predict the rockbursts. However, in hard coal mines, there are stroke or stress-stroke rockbursts in which the fracture of a thick layer of sandstone plays an essential role in predicting rockbursts. The occurrence of rockbursts in coal mines is complex, and their prediction is even more difficult than in other mines. In recent years, the interest in machine learning algorithms for solving complex nonlinear problems has increased, which also applies to geosciences. This study attempts to use machine learning algorithms, i.e. neural network, decision tree, random forest, gradient boosting, and extreme gradient boosting(XGB), to assess the rockburst hazard of an active hard coal mine in the Upper Silesian Coal Basin. The rock mass bursting tendency index WTGthat describes the tendency of the seam-surrounding rock system to rockbursts and the anomaly of the vertical stress component were applied for this purpose. Especially, the decision tree and neural network models were proved to be effective in correctly distinguishing rockbursts from tremors, after which the excavation was not damaged. On average, these models correctly classified about 80% of the rockbursts in the testing datasets.
基金Supported by the National Natural Science Foundation of China (60661003)the Research Project Department of Education of Jiangxi Province (GJJ10566)
文摘In this paper, we explore a novel ensemble method for spectral clustering. In contrast to the traditional clustering ensemble methods that combine all the obtained clustering results, we propose the adaptive spectral clustering ensemble method to achieve a better clustering solution. This method can adaptively assess the number of the component members, which is not owned by many other algorithms. The component clusterings of the ensemble system are generated by spectral clustering （SC） which bears some good characteristics to engender the diverse committees. The selection process works by evaluating the generated component spectral clustering through resampling technique and population-based incremental learning algorithm （PBIL）. Experimental results on UCI datasets demonstrate that the proposed algorithm can achieve better results compared with traditional clustering ensemble methods, especially when the number of component clusterings is large.
基金supported by the College of Agriculture,Shiraz University(Grant No.97GRC1M271143)funding from the UK Biotechnology and Biological Sciences Research Council(BBSRC)funded by BBSRC grant award BBS/E/C/000I0330–Soil to Nutrition project 3–Sustainable intensification:optimisation at multiple scales。
文摘This investigation assessed the efficacy of 10 widely used machine learning algorithms(MLA)comprising the least absolute shrinkage and selection operator(LASSO),generalized linear model(GLM),stepwise generalized linear model(SGLM),elastic net(ENET),partial least square(PLS),ridge regression,support vector machine(SVM),classification and regression trees(CART),bagged CART,and random forest(RF)for gully erosion susceptibility mapping(GESM)in Iran.The location of 462 previously existing gully erosion sites were mapped through widespread field investigations,of which 70%(323)and 30%(139)of observations were arbitrarily divided for algorithm calibration and validation.Twelve controlling factors for gully erosion,namely,soil texture,annual mean rainfall,digital elevation model(DEM),drainage density,slope,lithology,topographic wetness index(TWI),distance from rivers,aspect,distance from roads,plan curvature,and profile curvature were ranked in terms of their importance using each MLA.The MLA were compared using a training dataset for gully erosion and statistical measures such as RMSE(root mean square error),MAE(mean absolute error),and R-squared.Based on the comparisons among MLA,the RF algorithm exhibited the minimum RMSE and MAE and the maximum value of R-squared,and was therefore selected as the best model.The variable importance evaluation using the RF model revealed that distance from rivers had the highest significance in influencing the occurrence of gully erosion whereas plan curvature had the least importance.According to the GESM generated using RF,most of the study area is predicted to have a low(53.72%)or moderate(29.65%)susceptibility to gully erosion,whereas only a small area is identified to have a high(12.56%)or very high(4.07%)susceptibility.The outcome generated by RF model is validated using the ROC(Receiver Operating Characteristics)curve approach,which returned an area under the curve(AUC)of 0.985,proving the excellent forecasting ability of the model.The GESM prepared using the RF algorithm can aid decision-makers in targeting remedial actions for minimizing the damage caused by gully erosion.
基金the Doctorate Foundation of the Engineering College, Air Force Engineering University.
文摘The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are inferred based on the maximum likelihood rule. Secondly, the forgetting factor and learning rate factor are redefined, and their still more general formulations are obtained by analyzing their practical functions. Lastly, the convergence of the proposed algorithm is proved to enable the estimation converge to a local maximum of the data likelihood function according to the stochastic approximation theory. The experiments show that the proposed learning algorithm excels the formers both in converging rate and accuracy.
文摘Big data analytic techniques associated with machine learning algorithms are playing an increasingly important role in various application fields,including stock market investment.However,few studies have focused on forecasting daily stock market returns,especially when using powerful machine learning techniques,such as deep neural networks(DNNs),to perform the analyses.DNNs employ various deep learning algorithms based on the combination of network structure,activation function,and model parameters,with their performance depending on the format of the data representation.This paper presents a comprehensive big data analytics process to predict the daily return direction of the SPDR S&P 500 ETF(ticker symbol:SPY)based on 60 financial and economic features.DNNs and traditional artificial neural networks(ANNs)are then deployed over the entire preprocessed but untransformed dataset,along with two datasets transformed via principal component analysis(PCA),to predict the daily direction of future stock market index returns.While controlling for overfitting,a pattern for the classification accuracy of the DNNs is detected and demonstrated as the number of the hidden layers increases gradually from 12 to 1000.Moreover,a set of hypothesis testing procedures are implemented on the classification,and the simulation results show that the DNNs using two PCA-represented datasets give significantly higher classification accuracy than those using the entire untransformed dataset,as well as several other hybrid machine learning algorithms.In addition,the trading strategies guided by the DNN classification process based on PCA-represented data perform slightly better than the others tested,including in a comparison against two standard benchmarks.
