Due to the anonymity of blockchain,frequent security incidents and attacks occur through it,among which the Ponzi scheme smart contract is a classic type of fraud resulting in huge economic losses.Machine learningbase...Due to the anonymity of blockchain,frequent security incidents and attacks occur through it,among which the Ponzi scheme smart contract is a classic type of fraud resulting in huge economic losses.Machine learningbased methods are believed to be promising for detecting ethereum Ponzi schemes.However,there are still some flaws in current research,e.g.,insufficient feature extraction of Ponzi scheme smart contracts,without considering class imbalance.In addition,there is room for improvement in detection precision.Aiming at the above problems,this paper proposes an ethereum Ponzi scheme detection scheme through opcode context analysis and adaptive boosting(AdaBoost)algorithm.Firstly,this paper uses the n-gram algorithm to extract more comprehensive contract opcode features and combine them with contract account features,which helps to improve the feature extraction effect.Meanwhile,adaptive synthetic sampling(ADASYN)is introduced to deal with class imbalanced data,and integrated with the Adaboost classifier.Finally,this paper uses the improved AdaBoost classifier for the identification of Ponzi scheme contracts.Experimentally,this paper tests our model in real-world smart contracts and compares it with representative methods in the aspect of F1-score and precision.Moreover,this article compares and discusses the state of art methods with our method in four aspects:data acquisition,data preprocessing,feature extraction,and classifier design.Both experiment and discussion validate the effectiveness of our model.展开更多
HIV and AIDS has continued to be a major public health concern, and hence one of the epidemics that the world resolved to end by 2030 as highlighted in sustainable development goals (SDGs). A colossal amount of effort...HIV and AIDS has continued to be a major public health concern, and hence one of the epidemics that the world resolved to end by 2030 as highlighted in sustainable development goals (SDGs). A colossal amount of effort has been taken to reduce new HIV infections, but there are still a significant number of new infections reported. HIV prevalence is more skewed towards the key population who include female sex workers (FSW), men who have sex with men (MSM), and people who inject drugs (PWID). The study design was retrospective and focused on key population enrolled in a comprehensive HIV and AIDS programme by the Kenya Red Cross Society from July 2019 to June 2021. Individuals who were either lost to follow up, defaulted (dropped out, transferred out, or relocated) or died were classified as attrition;while those who were active and alive by the end of the study were classified as retention. The study used density analysis to determine the spatial differences of key population attrition in the 19 targeted counties, and used Kilifi county as an example to map attrition cases in smaller administrative areas (sub-county level). The study used synthetic minority oversampling technique-nominal continuous (SMOTE-NC) to balance the datasets since the cases of attrition were much less than retention. The random survival forests model was then fitted to the balanced dataset. The model correctly identified attrition cases using the predicted ensemble mortality and their survival time using the estimated Kaplan-Meier survival function. The predictive performance of the model was strong and way better than random chance with concordance indices greater than 0.75.展开更多
According to the oversampling imaging characteristics, an infrared small target detection method based on deep learning is proposed. A 7-layer deep convolutional neural network(CNN) is designed to automatically extrac...According to the oversampling imaging characteristics, an infrared small target detection method based on deep learning is proposed. A 7-layer deep convolutional neural network(CNN) is designed to automatically extract small target features and suppress clutters in an end-to-end manner. The input of CNN is an original oversampling image while the output is a cluttersuppressed feature map. The CNN contains only convolution and non-linear operations, and the resolution of the output feature map is the same as that of the input image. The L1-norm loss function is used, and a mass of training data is generated to train the network effectively. Results show that compared with several baseline methods, the proposed method improves the signal clutter ratio gain and background suppression factor by 3 – 4 orders of magnitude, and has more powerful target detection performance.展开更多
Learning from imbalanced data is one of the greatest challenging problems in binary classification,and this problem has gained more importance in recent years.When the class distribution is imbalanced,classical machin...Learning from imbalanced data is one of the greatest challenging problems in binary classification,and this problem has gained more importance in recent years.When the class distribution is imbalanced,classical machine learning algorithms tend to move strongly towards the majority class and disregard the minority.Therefore,the accuracy may be high,but the model cannot recognize data instances in the minority class to classify them,leading to many misclassifications.Different methods have been proposed in the literature to handle the imbalance problem,but most are complicated and tend to simulate unnecessary noise.In this paper,we propose a simple oversampling method based on Multivariate Gaussian distribution and K-means clustering,called GK-Means.The new method aims to avoid generating noise and control imbalances between and within classes.Various experiments have been carried out with six classifiers and four oversampling methods.Experimental results on different imbalanced datasets show that the proposed GK-Means outperforms other oversampling methods and improves classification performance as measured by F1-score and Accuracy.展开更多
An efficient single-carrier symbol synchronization method is proposed in this paper,which can work under a very low oversampling rate.This method is based on the frequency aliasing squared timing recovery assisted by ...An efficient single-carrier symbol synchronization method is proposed in this paper,which can work under a very low oversampling rate.This method is based on the frequency aliasing squared timing recovery assisted by pilot symbols and time domain filter.With frequency aliasing squared timing recovery with pilots,it is accessible to estimate timing error under oversampling rate less than 2.The time domain filter simultaneously performs matched-filtering and arbitrary interpolation.Because of pilot assisting,timing error estimation can be free from alias and self noise,so our method has good performance.Compared with traditional time-domain methods requiring oversampling rate above 2,this method can be adapted to any rational oversampling rate including less than 2.Moreover,compared with symbol synchronization in frequency domain which can operate under low oversampling rate,our method saves the complicated operation of conversion between time domain and frequency domain.By low oversampling rate and resource saving filter,this method is suitable for ultra-high-speed communication systems under resource-restricted hardware.The paper carries on the simulation and realization under 64QAM system.The simulation result shows that the loss is very low(less than 0.5 dB),and the real-time implementation on field programmable gate array(FPGA)also works fine.展开更多
Traditional ECG acquisition system lacks for flexibility. To improve the flexibility of ECG acquisition system and the signal-to-noise ratio of ECG, a new ECG acquisition system was designed based on DAQ card and Labv...Traditional ECG acquisition system lacks for flexibility. To improve the flexibility of ECG acquisition system and the signal-to-noise ratio of ECG, a new ECG acquisition system was designed based on DAQ card and Labview and oversampling was implemented in Labview. And analog signal conditioning circuit was improved on. The result indicated that the system could detect ECG signal accurately with high signal-to-noise ratio and the signal processing methods could be adjusted easily. So the new system can satisfy many kinds of ECG acquisition. It is a flexible experiment platform for exploring new ECG acquisition methods.展开更多
A reconfigurable intelligent surface(RIS)aided massive multiple-input multiple-output(MIMO)system is considered,where the base station employs a large antenna array with low-cost and low-power 1-bit analog-to-digital ...A reconfigurable intelligent surface(RIS)aided massive multiple-input multiple-output(MIMO)system is considered,where the base station employs a large antenna array with low-cost and low-power 1-bit analog-to-digital converters(ADCs).To compensate for the per-formance loss caused by the coarse quantization,oversampling is applied at the receiver.The main challenge for the acquisition of cascaded channel state information in such a system is to handle the distortion caused by the 1-bit quantization and the sample correlation caused by oversampling.In this work,Bussgang decomposition is applied to deal with the coarse quantization,and a Markov chain is developed to char-acterize the banded structure of the oversampling filter.An approximate message-passing based algorithm is proposed for the estimation of the cascaded channels.Simulation results demonstrate that our proposed 1-bit systems with oversampling can approach the 2-bit systems in terms of the mean square error performance while the former consumes much less power at the receiver.展开更多
Recently,anomaly detection(AD)in streaming data gained significant attention among research communities due to its applicability in finance,business,healthcare,education,etc.The recent developments of deep learning(DL...Recently,anomaly detection(AD)in streaming data gained significant attention among research communities due to its applicability in finance,business,healthcare,education,etc.The recent developments of deep learning(DL)models find helpful in the detection and classification of anomalies.This article designs an oversampling with an optimal deep learning-based streaming data classification(OS-ODLSDC)model.The aim of the OSODLSDC model is to recognize and classify the presence of anomalies in the streaming data.The proposed OS-ODLSDC model initially undergoes preprocessing step.Since streaming data is unbalanced,support vector machine(SVM)-Synthetic Minority Over-sampling Technique(SVM-SMOTE)is applied for oversampling process.Besides,the OS-ODLSDC model employs bidirectional long short-term memory(Bi LSTM)for AD and classification.Finally,the root means square propagation(RMSProp)optimizer is applied for optimal hyperparameter tuning of the Bi LSTM model.For ensuring the promising performance of the OS-ODLSDC model,a wide-ranging experimental analysis is performed using three benchmark datasets such as CICIDS 2018,KDD-Cup 1999,and NSL-KDD datasets.展开更多
BACKGROUND Postoperative delirium,particularly prevalent in elderly patients after abdominal cancer surgery,presents significant challenges in clinical management.AIM To develop a synthetic minority oversampling techn...BACKGROUND Postoperative delirium,particularly prevalent in elderly patients after abdominal cancer surgery,presents significant challenges in clinical management.AIM To develop a synthetic minority oversampling technique(SMOTE)-based model for predicting postoperative delirium in elderly abdominal cancer patients.METHODS In this retrospective cohort study,we analyzed data from 611 elderly patients who underwent abdominal malignant tumor surgery at our hospital between September 2020 and October 2022.The incidence of postoperative delirium was recorded for 7 d post-surgery.Patients were divided into delirium and non-delirium groups based on the occurrence of postoperative delirium or not.A multivariate logistic regression model was used to identify risk factors and develop a predictive model for postoperative delirium.The SMOTE technique was applied to enhance the model by oversampling the delirium cases.The model’s predictive accuracy was then validated.RESULTS In our study involving 611 elderly patients with abdominal malignant tumors,multivariate logistic regression analysis identified significant risk factors for postoperative delirium.These included the Charlson comorbidity index,American Society of Anesthesiologists classification,history of cerebrovascular disease,surgical duration,perioperative blood transfusion,and postoperative pain score.The incidence rate of postoperative delirium in our study was 22.91%.The original predictive model(P1)exhibited an area under the receiver operating characteristic curve of 0.862.In comparison,the SMOTE-based logistic early warning model(P2),which utilized the SMOTE oversampling algorithm,showed a slightly lower but comparable area under the curve of 0.856,suggesting no significant difference in performance between the two predictive approaches.CONCLUSION This study confirms that the SMOTE-enhanced predictive model for postoperative delirium in elderly abdominal tumor patients shows performance equivalent to that of traditional methods,effectively addressing data imbalance.展开更多
Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training dataset...Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training datasets limit the training performance and generalization.The model-driven deep learning introduces the diffraction model into the neural network.It eliminates the need for the labeled training dataset and has been extensively applied to hologram generation.However,the existing model-driven deep learning algorithms face the problem of insufficient constraints.In this study,we propose a model-driven neural network capable of high-fidelity 4K computer-generated hologram generation,called 4K Diffraction Model-driven Network(4K-DMDNet).The constraint of the reconstructed images in the frequency domain is strengthened.And a network structure that combines the residual method and sub-pixel convolution method is built,which effectively enhances the fitting ability of the network for inverse problems.The generalization of the 4K-DMDNet is demonstrated with binary,grayscale and 3D images.High-quality full-color optical reconstructions of the 4K holograms have been achieved at the wavelengths of 450 nm,520 nm,and 638 nm.展开更多
During tunnel boring machine(TBM)excavation,lithology identification is an important issue to understand tunnelling performance and avoid time-consuming excavation.However,site investigation generally lacks ground sam...During tunnel boring machine(TBM)excavation,lithology identification is an important issue to understand tunnelling performance and avoid time-consuming excavation.However,site investigation generally lacks ground samples and the information is subjective,heterogeneous,and imbalanced due to mixed ground conditions.In this study,an unsupervised(K-means)and synthetic minority oversampling technique(SMOTE)-guided light-gradient boosting machine(LightGBM)classifier is proposed to identify the soft ground tunnel classification and determine the imbalanced issue of tunnelling data.During the tunnel excavation,an earth pressure balance(EPB)TBM recorded 18 different operational parameters along with the three main tunnel lithologies.The proposed model is applied using Python low-code PyCaret library.Next,four decision tree-based classifiers were obtained in a short time period with automatic hyperparameter tuning to determine the best model for clustering-guided SMOTE application.In addition,the Shapley additive explanation(SHAP)was implemented to avoid the model black box problem.The proposed model was evaluated using different metrics such as accuracy,F1 score,precision,recall,and receiver operating characteristics(ROC)curve to obtain a reasonable outcome for the minority class.It shows that the proposed model can provide significant tunnel lithology identification based on the operational parameters of EPB-TBM.The proposed method can be applied to heterogeneous tunnel formations with several TBM operational parameters to describe the tunnel lithologies for efficient tunnelling.展开更多
Accurate regulation of two-dimensional materials has become an effective strategy to develop a wide range of catalytic applications.The introduction of heterogeneous components has a significant impact on the performa...Accurate regulation of two-dimensional materials has become an effective strategy to develop a wide range of catalytic applications.The introduction of heterogeneous components has a significant impact on the performance of materials,which makes it difficult to discover and understand the structure-property relationships at the atomic level.Here,we developed a novel and efficient ensemble learning classifier with synthetic minority oversampling technique(SMOTE) to discover all possible arsenene catalysts with implanted heteroatoms for hydrogen evolution reaction(HER).A total of 850 doped arsenenes were collected as a database and 140 modified arsenene materials with different doping atoms and doping sites were identified as promising candidate catalysts for HER,with a machine learning prediction accuracy of 81%.Based on the results of machine learning,we proposed 13 low-cost and easily synthesized two-dimensional Fe-doped arsenene catalytic materials that are expected to contribute to high-efficient HER.The proposed ensemble method achieved high prediction accuracy,but millions of times faster to predict Gibbs free energies and only required a small amount of data.This study indicates that the presented ensemble learning classifier is capable of screening high-efficient catalysts,and can be further extended to predict other two-dimensional catalysts with delicate regulation.展开更多
Alzheimer’s disease(AD)is a neurodevelopmental impairment that results in a person’s behavior,thinking,and memory loss.Themost common symptoms ofADare losingmemory and early aging.In addition to these,there are seve...Alzheimer’s disease(AD)is a neurodevelopmental impairment that results in a person’s behavior,thinking,and memory loss.Themost common symptoms ofADare losingmemory and early aging.In addition to these,there are several serious impacts ofAD.However,the impact ofADcanbemitigatedby early-stagedetection though it cannot be cured permanently.Early-stage detection is the most challenging task for controlling and mitigating the impact of AD.The study proposes a predictive model to detect AD in the initial phase based on machine learning and a deep learning approach to address the issue.To build a predictive model,open-source data was collected where five stages of images of AD were available as Cognitive Normal(CN),Early Mild Cognitive Impairment(EMCI),Mild Cognitive Impairment(MCI),Late Mild Cognitive Impairment(LMCI),and AD.Every stage of AD is considered as a class,and then the dataset was divided into three parts binary class,three class,and five class.In this research,we applied different preprocessing steps with augmentation techniques to efficiently identifyAD.It integrates a random oversampling technique to handle the imbalance problem from target classes,mitigating the model overfitting and biases.Then three machine learning classifiers,such as random forest(RF),K-Nearest neighbor(KNN),and support vector machine(SVM),and two deep learning methods,such as convolutional neuronal network(CNN)and artificial neural network(ANN)were applied on these datasets.After analyzing the performance of the used models and the datasets,it is found that CNN with binary class outperformed 88.20%accuracy.The result of the study indicates that the model is highly potential to detect AD in the initial phase.展开更多
Most modern technologies,such as social media,smart cities,and the internet of things(IoT),rely on big data.When big data is used in the real-world applications,two data challenges such as class overlap and class imba...Most modern technologies,such as social media,smart cities,and the internet of things(IoT),rely on big data.When big data is used in the real-world applications,two data challenges such as class overlap and class imbalance arises.When dealing with large datasets,most traditional classifiers are stuck in the local optimum problem.As a result,it’s necessary to look into new methods for dealing with large data collections.Several solutions have been proposed for overcoming this issue.The rapid growth of the available data threatens to limit the usefulness of many traditional methods.Methods such as oversampling and undersampling have shown great promises in addressing the issues of class imbalance.Among all of these techniques,Synthetic Minority Oversampling TechniquE(SMOTE)has produced the best results by generating synthetic samples for the minority class in creating a balanced dataset.The issue is that their practical applicability is restricted to problems involving tens of thousands or lower instances of each.In this paper,we have proposed a parallel mode method using SMOTE and MapReduce strategy,this distributes the operation of the algorithm among a group of computational nodes for addressing the aforementioned problem.Our proposed solution has been divided into three stages.Thefirst stage involves the process of splitting the data into different blocks using a mapping function,followed by a pre-processing step for each mapping block that employs a hybrid SMOTE algo-rithm for solving the class imbalanced problem.On each map block,a decision tree model would be constructed.Finally,the decision tree blocks would be com-bined for creating a classification model.We have used numerous datasets with up to 4 million instances in our experiments for testing the proposed scheme’s cap-abilities.As a result,the Hybrid SMOTE appears to have good scalability within the framework proposed,and it also cuts down the processing time.展开更多
Twitter has emerged as a platform that produces new data every day through its users which can be utilized for various purposes.People express their unique ideas and views onmultiple topics thus providing vast knowled...Twitter has emerged as a platform that produces new data every day through its users which can be utilized for various purposes.People express their unique ideas and views onmultiple topics thus providing vast knowledge.Sentiment analysis is critical from the corporate and political perspectives as it can impact decision-making.Since the proliferation of COVID-19,it has become an important challenge to detect the sentiment of COVID-19-related tweets so that people’s opinions can be tracked.The purpose of this research is to detect the sentiment of people regarding this problem with limited data as it can be challenging considering the various textual characteristics that must be analyzed.Hence,this research presents a deep learning-based model that utilizes the positives of random minority oversampling combined with class label analysis to achieve the best results for sentiment analysis.This research specifically focuses on utilizing class label analysis to deal with the multiclass problem by combining the class labels with a similar overall sentiment.This can be particularly helpful when dealing with smaller datasets.Furthermore,our proposed model integrates various preprocessing steps with random minority oversampling and various deep learning algorithms including standard deep learning and bi-directional deep learning algorithms.This research explores several algorithms and their impact on sentiment analysis tasks and concludes that bidirectional neural networks do not provide any advantage over standard neural networks as standard Neural Networks provide slightly better results than their bidirectional counterparts.The experimental results validate that our model offers excellent results with a validation accuracy of 92.5%and an F1 measure of 0.92.展开更多
This paper aims to build an employee attrition classification model based on the Stacking algorithm.Oversampling algorithm is applied to address the issue of data imbalance and the Randomforest feature importance rank...This paper aims to build an employee attrition classification model based on the Stacking algorithm.Oversampling algorithm is applied to address the issue of data imbalance and the Randomforest feature importance ranking method is used to resolve the overfitting problem after data cleaning and preprocessing.Then,different algorithms are used to establish classification models as control experiments,and R-squared indicators are used to compare.Finally,the Stacking algorithm is used to establish the final classification model.This model has practical and significant implications for both human resource management and employee attrition analysis.展开更多
This study aims to develop a low-cost refractometer for measuring the sucrose content of fruit juice,which is an important factor affecting human health.While laboratory-grade refractometers are expensive and unsuitab...This study aims to develop a low-cost refractometer for measuring the sucrose content of fruit juice,which is an important factor affecting human health.While laboratory-grade refractometers are expensive and unsuitable for personal use,existing low-cost commercial options lack stability and accuracy.To address this gap,we propose a refractometer that replaces the expensive CCD sensor and light source with a conventional LED and a reasonably priced CMOS sensor.By analyzing the output waveform pattern of the CMOS sensor,we achieve high precision with a personal-use-appropriate accuracy of 0.1%.We tested the proposed refractometer by conducting 100 repeated measurements on various fruit juice samples,and the results demonstrate its reliability and consistency.Running on a 48 MHz ARM processor,the algorithm can acquire data within 0.2 seconds.Our low-cost refractometer is suitable for personal health management and small-scale production,providing an affordable and reliable method for measuring sucrose concentration in fruit juice.It improves upon the existing low-cost options by offering better stability and accuracy.This accessible tool has potential applications in optimizing the sucrose content of fruit juice for better health and quality control.展开更多
Real-time prediction of the rock mass class in front of the tunnel face is essential for the adaptive adjustment of tunnel boring machines(TBMs).During the TBM tunnelling process,a large number of operation data are g...Real-time prediction of the rock mass class in front of the tunnel face is essential for the adaptive adjustment of tunnel boring machines(TBMs).During the TBM tunnelling process,a large number of operation data are generated,reflecting the interaction between the TBM system and surrounding rock,and these data can be used to evaluate the rock mass quality.This study proposed a stacking ensemble classifier for the real-time prediction of the rock mass classification using TBM operation data.Based on the Songhua River water conveyance project,a total of 7538 TBM tunnelling cycles and the corresponding rock mass classes are obtained after data preprocessing.Then,through the tree-based feature selection method,10 key TBM operation parameters are selected,and the mean values of the 10 selected features in the stable phase after removing outliers are calculated as the inputs of classifiers.The preprocessed data are randomly divided into the training set(90%)and test set(10%)using simple random sampling.Besides stacking ensemble classifier,seven individual classifiers are established as the comparison.These classifiers include support vector machine(SVM),k-nearest neighbors(KNN),random forest(RF),gradient boosting decision tree(GBDT),decision tree(DT),logistic regression(LR)and multilayer perceptron(MLP),where the hyper-parameters of each classifier are optimised using the grid search method.The prediction results show that the stacking ensemble classifier has a better performance than individual classifiers,and it shows a more powerful learning and generalisation ability for small and imbalanced samples.Additionally,a relative balance training set is obtained by the synthetic minority oversampling technique(SMOTE),and the influence of sample imbalance on the prediction performance is discussed.展开更多
基金This work was supported by National Key R&D Program of China(Grant Numbers 2020YFB1005900,2022YFB3305802).
文摘Due to the anonymity of blockchain,frequent security incidents and attacks occur through it,among which the Ponzi scheme smart contract is a classic type of fraud resulting in huge economic losses.Machine learningbased methods are believed to be promising for detecting ethereum Ponzi schemes.However,there are still some flaws in current research,e.g.,insufficient feature extraction of Ponzi scheme smart contracts,without considering class imbalance.In addition,there is room for improvement in detection precision.Aiming at the above problems,this paper proposes an ethereum Ponzi scheme detection scheme through opcode context analysis and adaptive boosting(AdaBoost)algorithm.Firstly,this paper uses the n-gram algorithm to extract more comprehensive contract opcode features and combine them with contract account features,which helps to improve the feature extraction effect.Meanwhile,adaptive synthetic sampling(ADASYN)is introduced to deal with class imbalanced data,and integrated with the Adaboost classifier.Finally,this paper uses the improved AdaBoost classifier for the identification of Ponzi scheme contracts.Experimentally,this paper tests our model in real-world smart contracts and compares it with representative methods in the aspect of F1-score and precision.Moreover,this article compares and discusses the state of art methods with our method in four aspects:data acquisition,data preprocessing,feature extraction,and classifier design.Both experiment and discussion validate the effectiveness of our model.
文摘HIV and AIDS has continued to be a major public health concern, and hence one of the epidemics that the world resolved to end by 2030 as highlighted in sustainable development goals (SDGs). A colossal amount of effort has been taken to reduce new HIV infections, but there are still a significant number of new infections reported. HIV prevalence is more skewed towards the key population who include female sex workers (FSW), men who have sex with men (MSM), and people who inject drugs (PWID). The study design was retrospective and focused on key population enrolled in a comprehensive HIV and AIDS programme by the Kenya Red Cross Society from July 2019 to June 2021. Individuals who were either lost to follow up, defaulted (dropped out, transferred out, or relocated) or died were classified as attrition;while those who were active and alive by the end of the study were classified as retention. The study used density analysis to determine the spatial differences of key population attrition in the 19 targeted counties, and used Kilifi county as an example to map attrition cases in smaller administrative areas (sub-county level). The study used synthetic minority oversampling technique-nominal continuous (SMOTE-NC) to balance the datasets since the cases of attrition were much less than retention. The random survival forests model was then fitted to the balanced dataset. The model correctly identified attrition cases using the predicted ensemble mortality and their survival time using the estimated Kaplan-Meier survival function. The predictive performance of the model was strong and way better than random chance with concordance indices greater than 0.75.
基金supported by the National Key Research and Development Program of China(2016YFB0500901)the Natural Science Foundation of Shanghai(18ZR1437200)the Satellite Mapping Technology and Application National Key Laboratory of Geographical Information Bureau(KLSMTA-201709)
文摘According to the oversampling imaging characteristics, an infrared small target detection method based on deep learning is proposed. A 7-layer deep convolutional neural network(CNN) is designed to automatically extract small target features and suppress clutters in an end-to-end manner. The input of CNN is an original oversampling image while the output is a cluttersuppressed feature map. The CNN contains only convolution and non-linear operations, and the resolution of the output feature map is the same as that of the input image. The L1-norm loss function is used, and a mass of training data is generated to train the network effectively. Results show that compared with several baseline methods, the proposed method improves the signal clutter ratio gain and background suppression factor by 3 – 4 orders of magnitude, and has more powerful target detection performance.
文摘Learning from imbalanced data is one of the greatest challenging problems in binary classification,and this problem has gained more importance in recent years.When the class distribution is imbalanced,classical machine learning algorithms tend to move strongly towards the majority class and disregard the minority.Therefore,the accuracy may be high,but the model cannot recognize data instances in the minority class to classify them,leading to many misclassifications.Different methods have been proposed in the literature to handle the imbalance problem,but most are complicated and tend to simulate unnecessary noise.In this paper,we propose a simple oversampling method based on Multivariate Gaussian distribution and K-means clustering,called GK-Means.The new method aims to avoid generating noise and control imbalances between and within classes.Various experiments have been carried out with six classifiers and four oversampling methods.Experimental results on different imbalanced datasets show that the proposed GK-Means outperforms other oversampling methods and improves classification performance as measured by F1-score and Accuracy.
文摘An efficient single-carrier symbol synchronization method is proposed in this paper,which can work under a very low oversampling rate.This method is based on the frequency aliasing squared timing recovery assisted by pilot symbols and time domain filter.With frequency aliasing squared timing recovery with pilots,it is accessible to estimate timing error under oversampling rate less than 2.The time domain filter simultaneously performs matched-filtering and arbitrary interpolation.Because of pilot assisting,timing error estimation can be free from alias and self noise,so our method has good performance.Compared with traditional time-domain methods requiring oversampling rate above 2,this method can be adapted to any rational oversampling rate including less than 2.Moreover,compared with symbol synchronization in frequency domain which can operate under low oversampling rate,our method saves the complicated operation of conversion between time domain and frequency domain.By low oversampling rate and resource saving filter,this method is suitable for ultra-high-speed communication systems under resource-restricted hardware.The paper carries on the simulation and realization under 64QAM system.The simulation result shows that the loss is very low(less than 0.5 dB),and the real-time implementation on field programmable gate array(FPGA)also works fine.
文摘Traditional ECG acquisition system lacks for flexibility. To improve the flexibility of ECG acquisition system and the signal-to-noise ratio of ECG, a new ECG acquisition system was designed based on DAQ card and Labview and oversampling was implemented in Labview. And analog signal conditioning circuit was improved on. The result indicated that the system could detect ECG signal accurately with high signal-to-noise ratio and the signal processing methods could be adjusted easily. So the new system can satisfy many kinds of ECG acquisition. It is a flexible experiment platform for exploring new ECG acquisition methods.
文摘A reconfigurable intelligent surface(RIS)aided massive multiple-input multiple-output(MIMO)system is considered,where the base station employs a large antenna array with low-cost and low-power 1-bit analog-to-digital converters(ADCs).To compensate for the per-formance loss caused by the coarse quantization,oversampling is applied at the receiver.The main challenge for the acquisition of cascaded channel state information in such a system is to handle the distortion caused by the 1-bit quantization and the sample correlation caused by oversampling.In this work,Bussgang decomposition is applied to deal with the coarse quantization,and a Markov chain is developed to char-acterize the banded structure of the oversampling filter.An approximate message-passing based algorithm is proposed for the estimation of the cascaded channels.Simulation results demonstrate that our proposed 1-bit systems with oversampling can approach the 2-bit systems in terms of the mean square error performance while the former consumes much less power at the receiver.
文摘Recently,anomaly detection(AD)in streaming data gained significant attention among research communities due to its applicability in finance,business,healthcare,education,etc.The recent developments of deep learning(DL)models find helpful in the detection and classification of anomalies.This article designs an oversampling with an optimal deep learning-based streaming data classification(OS-ODLSDC)model.The aim of the OSODLSDC model is to recognize and classify the presence of anomalies in the streaming data.The proposed OS-ODLSDC model initially undergoes preprocessing step.Since streaming data is unbalanced,support vector machine(SVM)-Synthetic Minority Over-sampling Technique(SVM-SMOTE)is applied for oversampling process.Besides,the OS-ODLSDC model employs bidirectional long short-term memory(Bi LSTM)for AD and classification.Finally,the root means square propagation(RMSProp)optimizer is applied for optimal hyperparameter tuning of the Bi LSTM model.For ensuring the promising performance of the OS-ODLSDC model,a wide-ranging experimental analysis is performed using three benchmark datasets such as CICIDS 2018,KDD-Cup 1999,and NSL-KDD datasets.
基金Supported by Discipline Advancement Program of Shanghai Fourth People’s Hospital,No.SY-XKZT-2020-2013.
文摘BACKGROUND Postoperative delirium,particularly prevalent in elderly patients after abdominal cancer surgery,presents significant challenges in clinical management.AIM To develop a synthetic minority oversampling technique(SMOTE)-based model for predicting postoperative delirium in elderly abdominal cancer patients.METHODS In this retrospective cohort study,we analyzed data from 611 elderly patients who underwent abdominal malignant tumor surgery at our hospital between September 2020 and October 2022.The incidence of postoperative delirium was recorded for 7 d post-surgery.Patients were divided into delirium and non-delirium groups based on the occurrence of postoperative delirium or not.A multivariate logistic regression model was used to identify risk factors and develop a predictive model for postoperative delirium.The SMOTE technique was applied to enhance the model by oversampling the delirium cases.The model’s predictive accuracy was then validated.RESULTS In our study involving 611 elderly patients with abdominal malignant tumors,multivariate logistic regression analysis identified significant risk factors for postoperative delirium.These included the Charlson comorbidity index,American Society of Anesthesiologists classification,history of cerebrovascular disease,surgical duration,perioperative blood transfusion,and postoperative pain score.The incidence rate of postoperative delirium in our study was 22.91%.The original predictive model(P1)exhibited an area under the receiver operating characteristic curve of 0.862.In comparison,the SMOTE-based logistic early warning model(P2),which utilized the SMOTE oversampling algorithm,showed a slightly lower but comparable area under the curve of 0.856,suggesting no significant difference in performance between the two predictive approaches.CONCLUSION This study confirms that the SMOTE-enhanced predictive model for postoperative delirium in elderly abdominal tumor patients shows performance equivalent to that of traditional methods,effectively addressing data imbalance.
基金We are grateful for financial supports from National Natural Science Foundation of China(62035003,61775117)China Postdoctoral Science Foundation(BX2021140)Tsinghua University Initiative Scientific Research Program(20193080075).
文摘Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training datasets limit the training performance and generalization.The model-driven deep learning introduces the diffraction model into the neural network.It eliminates the need for the labeled training dataset and has been extensively applied to hologram generation.However,the existing model-driven deep learning algorithms face the problem of insufficient constraints.In this study,we propose a model-driven neural network capable of high-fidelity 4K computer-generated hologram generation,called 4K Diffraction Model-driven Network(4K-DMDNet).The constraint of the reconstructed images in the frequency domain is strengthened.And a network structure that combines the residual method and sub-pixel convolution method is built,which effectively enhances the fitting ability of the network for inverse problems.The generalization of the 4K-DMDNet is demonstrated with binary,grayscale and 3D images.High-quality full-color optical reconstructions of the 4K holograms have been achieved at the wavelengths of 450 nm,520 nm,and 638 nm.
基金supported by Japan Society for the Promotion of Science KAKENHI(Grant No.JP22H01580).
文摘During tunnel boring machine(TBM)excavation,lithology identification is an important issue to understand tunnelling performance and avoid time-consuming excavation.However,site investigation generally lacks ground samples and the information is subjective,heterogeneous,and imbalanced due to mixed ground conditions.In this study,an unsupervised(K-means)and synthetic minority oversampling technique(SMOTE)-guided light-gradient boosting machine(LightGBM)classifier is proposed to identify the soft ground tunnel classification and determine the imbalanced issue of tunnelling data.During the tunnel excavation,an earth pressure balance(EPB)TBM recorded 18 different operational parameters along with the three main tunnel lithologies.The proposed model is applied using Python low-code PyCaret library.Next,four decision tree-based classifiers were obtained in a short time period with automatic hyperparameter tuning to determine the best model for clustering-guided SMOTE application.In addition,the Shapley additive explanation(SHAP)was implemented to avoid the model black box problem.The proposed model was evaluated using different metrics such as accuracy,F1 score,precision,recall,and receiver operating characteristics(ROC)curve to obtain a reasonable outcome for the minority class.It shows that the proposed model can provide significant tunnel lithology identification based on the operational parameters of EPB-TBM.The proposed method can be applied to heterogeneous tunnel formations with several TBM operational parameters to describe the tunnel lithologies for efficient tunnelling.
基金supported by the National Key R&D Program of China(No.2021YFC2100100)the National Natural Science Foundation of China(No.21901157)the Shanghai Science and Technology Project(No.21JC1403400)。
文摘Accurate regulation of two-dimensional materials has become an effective strategy to develop a wide range of catalytic applications.The introduction of heterogeneous components has a significant impact on the performance of materials,which makes it difficult to discover and understand the structure-property relationships at the atomic level.Here,we developed a novel and efficient ensemble learning classifier with synthetic minority oversampling technique(SMOTE) to discover all possible arsenene catalysts with implanted heteroatoms for hydrogen evolution reaction(HER).A total of 850 doped arsenenes were collected as a database and 140 modified arsenene materials with different doping atoms and doping sites were identified as promising candidate catalysts for HER,with a machine learning prediction accuracy of 81%.Based on the results of machine learning,we proposed 13 low-cost and easily synthesized two-dimensional Fe-doped arsenene catalytic materials that are expected to contribute to high-efficient HER.The proposed ensemble method achieved high prediction accuracy,but millions of times faster to predict Gibbs free energies and only required a small amount of data.This study indicates that the presented ensemble learning classifier is capable of screening high-efficient catalysts,and can be further extended to predict other two-dimensional catalysts with delicate regulation.
基金funded in part by the Natural Sciences and Engineering Research Council of Canada(NSERC)through Project Number:IFP22UQU4170008DSR0056.
文摘Alzheimer’s disease(AD)is a neurodevelopmental impairment that results in a person’s behavior,thinking,and memory loss.Themost common symptoms ofADare losingmemory and early aging.In addition to these,there are several serious impacts ofAD.However,the impact ofADcanbemitigatedby early-stagedetection though it cannot be cured permanently.Early-stage detection is the most challenging task for controlling and mitigating the impact of AD.The study proposes a predictive model to detect AD in the initial phase based on machine learning and a deep learning approach to address the issue.To build a predictive model,open-source data was collected where five stages of images of AD were available as Cognitive Normal(CN),Early Mild Cognitive Impairment(EMCI),Mild Cognitive Impairment(MCI),Late Mild Cognitive Impairment(LMCI),and AD.Every stage of AD is considered as a class,and then the dataset was divided into three parts binary class,three class,and five class.In this research,we applied different preprocessing steps with augmentation techniques to efficiently identifyAD.It integrates a random oversampling technique to handle the imbalance problem from target classes,mitigating the model overfitting and biases.Then three machine learning classifiers,such as random forest(RF),K-Nearest neighbor(KNN),and support vector machine(SVM),and two deep learning methods,such as convolutional neuronal network(CNN)and artificial neural network(ANN)were applied on these datasets.After analyzing the performance of the used models and the datasets,it is found that CNN with binary class outperformed 88.20%accuracy.The result of the study indicates that the model is highly potential to detect AD in the initial phase.
文摘Most modern technologies,such as social media,smart cities,and the internet of things(IoT),rely on big data.When big data is used in the real-world applications,two data challenges such as class overlap and class imbalance arises.When dealing with large datasets,most traditional classifiers are stuck in the local optimum problem.As a result,it’s necessary to look into new methods for dealing with large data collections.Several solutions have been proposed for overcoming this issue.The rapid growth of the available data threatens to limit the usefulness of many traditional methods.Methods such as oversampling and undersampling have shown great promises in addressing the issues of class imbalance.Among all of these techniques,Synthetic Minority Oversampling TechniquE(SMOTE)has produced the best results by generating synthetic samples for the minority class in creating a balanced dataset.The issue is that their practical applicability is restricted to problems involving tens of thousands or lower instances of each.In this paper,we have proposed a parallel mode method using SMOTE and MapReduce strategy,this distributes the operation of the algorithm among a group of computational nodes for addressing the aforementioned problem.Our proposed solution has been divided into three stages.Thefirst stage involves the process of splitting the data into different blocks using a mapping function,followed by a pre-processing step for each mapping block that employs a hybrid SMOTE algo-rithm for solving the class imbalanced problem.On each map block,a decision tree model would be constructed.Finally,the decision tree blocks would be com-bined for creating a classification model.We have used numerous datasets with up to 4 million instances in our experiments for testing the proposed scheme’s cap-abilities.As a result,the Hybrid SMOTE appears to have good scalability within the framework proposed,and it also cuts down the processing time.
基金This work was funded by the Deanship of Scientific Research at Jouf University under Grant Number(DSR2022-RG-0105).
文摘Twitter has emerged as a platform that produces new data every day through its users which can be utilized for various purposes.People express their unique ideas and views onmultiple topics thus providing vast knowledge.Sentiment analysis is critical from the corporate and political perspectives as it can impact decision-making.Since the proliferation of COVID-19,it has become an important challenge to detect the sentiment of COVID-19-related tweets so that people’s opinions can be tracked.The purpose of this research is to detect the sentiment of people regarding this problem with limited data as it can be challenging considering the various textual characteristics that must be analyzed.Hence,this research presents a deep learning-based model that utilizes the positives of random minority oversampling combined with class label analysis to achieve the best results for sentiment analysis.This research specifically focuses on utilizing class label analysis to deal with the multiclass problem by combining the class labels with a similar overall sentiment.This can be particularly helpful when dealing with smaller datasets.Furthermore,our proposed model integrates various preprocessing steps with random minority oversampling and various deep learning algorithms including standard deep learning and bi-directional deep learning algorithms.This research explores several algorithms and their impact on sentiment analysis tasks and concludes that bidirectional neural networks do not provide any advantage over standard neural networks as standard Neural Networks provide slightly better results than their bidirectional counterparts.The experimental results validate that our model offers excellent results with a validation accuracy of 92.5%and an F1 measure of 0.92.
文摘This paper aims to build an employee attrition classification model based on the Stacking algorithm.Oversampling algorithm is applied to address the issue of data imbalance and the Randomforest feature importance ranking method is used to resolve the overfitting problem after data cleaning and preprocessing.Then,different algorithms are used to establish classification models as control experiments,and R-squared indicators are used to compare.Finally,the Stacking algorithm is used to establish the final classification model.This model has practical and significant implications for both human resource management and employee attrition analysis.
文摘This study aims to develop a low-cost refractometer for measuring the sucrose content of fruit juice,which is an important factor affecting human health.While laboratory-grade refractometers are expensive and unsuitable for personal use,existing low-cost commercial options lack stability and accuracy.To address this gap,we propose a refractometer that replaces the expensive CCD sensor and light source with a conventional LED and a reasonably priced CMOS sensor.By analyzing the output waveform pattern of the CMOS sensor,we achieve high precision with a personal-use-appropriate accuracy of 0.1%.We tested the proposed refractometer by conducting 100 repeated measurements on various fruit juice samples,and the results demonstrate its reliability and consistency.Running on a 48 MHz ARM processor,the algorithm can acquire data within 0.2 seconds.Our low-cost refractometer is suitable for personal health management and small-scale production,providing an affordable and reliable method for measuring sucrose concentration in fruit juice.It improves upon the existing low-cost options by offering better stability and accuracy.This accessible tool has potential applications in optimizing the sucrose content of fruit juice for better health and quality control.
基金funded by the National Natural Science Foundation of China(Grant No.41941019)the State Key Laboratory of Hydroscience and Engineering(Grant No.2019-KY-03)。
文摘Real-time prediction of the rock mass class in front of the tunnel face is essential for the adaptive adjustment of tunnel boring machines(TBMs).During the TBM tunnelling process,a large number of operation data are generated,reflecting the interaction between the TBM system and surrounding rock,and these data can be used to evaluate the rock mass quality.This study proposed a stacking ensemble classifier for the real-time prediction of the rock mass classification using TBM operation data.Based on the Songhua River water conveyance project,a total of 7538 TBM tunnelling cycles and the corresponding rock mass classes are obtained after data preprocessing.Then,through the tree-based feature selection method,10 key TBM operation parameters are selected,and the mean values of the 10 selected features in the stable phase after removing outliers are calculated as the inputs of classifiers.The preprocessed data are randomly divided into the training set(90%)and test set(10%)using simple random sampling.Besides stacking ensemble classifier,seven individual classifiers are established as the comparison.These classifiers include support vector machine(SVM),k-nearest neighbors(KNN),random forest(RF),gradient boosting decision tree(GBDT),decision tree(DT),logistic regression(LR)and multilayer perceptron(MLP),where the hyper-parameters of each classifier are optimised using the grid search method.The prediction results show that the stacking ensemble classifier has a better performance than individual classifiers,and it shows a more powerful learning and generalisation ability for small and imbalanced samples.Additionally,a relative balance training set is obtained by the synthetic minority oversampling technique(SMOTE),and the influence of sample imbalance on the prediction performance is discussed.
