One-class classification problem has become a popular problem in many fields, with a wide range of applications in anomaly detection, fault diagnosis, and face recognition. We investigate the one-class classification ...One-class classification problem has become a popular problem in many fields, with a wide range of applications in anomaly detection, fault diagnosis, and face recognition. We investigate the one-class classification problem for second-order tensor data. Traditional vector-based one-class classification methods such as one-class support vector machine (OCSVM) and least squares one-class support vector machine (LSOCSVM) have limitations when tensor is used as input data, so we propose a new tensor one-class classification method, LSOCSTM, which directly uses tensor as input data. On one hand, using tensor as input data not only enables to classify tensor data, but also for vector data, classifying it after high dimensionalizing it into tensor still improves the classification accuracy and overcomes the over-fitting problem. On the other hand, different from one-class support tensor machine (OCSTM), we use squared loss instead of the original loss function so that we solve a series of linear equations instead of quadratic programming problems. Therefore, we use the distance to the hyperplane as a metric for classification, and the proposed method is more accurate and faster compared to existing methods. The experimental results show the high efficiency of the proposed method compared with several state-of-the-art methods.展开更多
BACKGROUND The relation between orthodontic treatment and temporomandibular disorders(TMDs)is under debate;the management of TMD during orthodontic treatment has always been a challenge.If TMD symptoms occur during or...BACKGROUND The relation between orthodontic treatment and temporomandibular disorders(TMDs)is under debate;the management of TMD during orthodontic treatment has always been a challenge.If TMD symptoms occur during orthodontic treatment,an immediate pause of orthodontic adjustments is recommended;the treatment can resume when the symptoms are managed and stabilized.CASE SUMMARY This case report presents a patient(26-year-old,female)with angle class I,skeletal class II and TMDs.The treatment was a hybrid of clear aligners,fixed appliances and temporary anchorage devices(TADs).After 3 mo resting and treatment on her TMD,the patient’s TMD symptom alleviated,but her anterior occlusion displayed deep overbite.Therefore,the fixed appliances with TAD were used to correct the anterior deep-bite and level maxillary and mandibular deep curves.After the levelling,the patient showed dual bite with centric relation and maximum intercuspation discrepancy on her occlusion.After careful examination of temporomandibular joints(TMJ)position,the stable bite splint and Invisible Mandibular Advancement appliance were used to reconstruct her occlusion.Eventually,the improved facial appearance and relatively stable occlusion were achieved.The 1-year follow-up records showed there was no obvious change in TMJ morphology,and her occlusion was stable.CONCLUSION TMD screening and monitoring is of great clinical importance in the TMD susceptible patients.Hybrid treatment with clear aligners and fixed appliances and TADs is an effective treatment modality for the complex cases.展开更多
This study analyzed the difference between using a downward breaststroke kick and a horizontal breaststroke kick in a sample of world class elite swimmers.We compared average muscle activity of the gluteus maximus,qua...This study analyzed the difference between using a downward breaststroke kick and a horizontal breaststroke kick in a sample of world class elite swimmers.We compared average muscle activity of the gluteus maximus,quadriceps femoris(vastus medialis and rectus femoris),hamstring/long head of the biceps femoris,gastrocnemius medialis,rectus abdominal,and erector spinae when using the downward breaststroke kick technique.We find that when this sample of swimmers utilized the downward breaststroke kick,max speed and velocity per stroke increased,measured by 12,788 EMG samples,where the results are highly correlated to duration of the aerodynamic buoyant force in breaststroke kick technique.The increases in performance observed from measuring the world class elite swimmers is highly correlated to the duration of the kick aerodynamic buoyant force.Among this sample of elite swimmers,the longer a swimmer demonstrates a buoyant force breaststroke kick,the lower the time in a 100 breaststroke.展开更多
Every application in a smart city environment like the smart grid,health monitoring, security, and surveillance generates non-stationary datastreams. Due to such nature, the statistical properties of data changes over...Every application in a smart city environment like the smart grid,health monitoring, security, and surveillance generates non-stationary datastreams. Due to such nature, the statistical properties of data changes overtime, leading to class imbalance and concept drift issues. Both these issuescause model performance degradation. Most of the current work has beenfocused on developing an ensemble strategy by training a new classifier on thelatest data to resolve the issue. These techniques suffer while training the newclassifier if the data is imbalanced. Also, the class imbalance ratio may changegreatly from one input stream to another, making the problem more complex.The existing solutions proposed for addressing the combined issue of classimbalance and concept drift are lacking in understating of correlation of oneproblem with the other. This work studies the association between conceptdrift and class imbalance ratio and then demonstrates how changes in classimbalance ratio along with concept drift affect the classifier’s performance.We analyzed the effect of both the issues on minority and majority classesindividually. To do this, we conducted experiments on benchmark datasetsusing state-of-the-art classifiers especially designed for data stream classification.Precision, recall, F1 score, and geometric mean were used to measure theperformance. Our findings show that when both class imbalance and conceptdrift problems occur together the performance can decrease up to 15%. Ourresults also show that the increase in the imbalance ratio can cause a 10% to15% decrease in the precision scores of both minority and majority classes.The study findings may help in designing intelligent and adaptive solutionsthat can cope with the challenges of non-stationary data streams like conceptdrift and class imbalance.展开更多
The development of image classification is one of the most important research topics in remote sensing. The prediction accuracy depends not only on the appropriate choice of the machine learning method but also on the...The development of image classification is one of the most important research topics in remote sensing. The prediction accuracy depends not only on the appropriate choice of the machine learning method but also on the quality of the training datasets. However, real-world data is not perfect and often suffers from noise. This paper gives an overview of noise filtering methods. Firstly, the types of noise and the consequences of class noise on machine learning are presented. Secondly, class noise handling methods at both the data level and the algorithm level are introduced. Then ensemble-based class noise handling methods including class noise removal, correction, and noise robust ensemble learners are presented. Finally, a summary of existing data-cleaning techniques is given.展开更多
From the National Agency for the Organization and Construction of Infrastructures (ANOCI) program to the Emerging Senegal Plan (PSE), construction in Senegal has improved considerably. However, deficiencies remain in ...From the National Agency for the Organization and Construction of Infrastructures (ANOCI) program to the Emerging Senegal Plan (PSE), construction in Senegal has improved considerably. However, deficiencies remain in the specification of materials used mainly in hydraulic concrete. They are generally related to the specification of aggregates for alkali reactivity and the choice of exposure classes specified in NF EN 206-1. The purpose of this article is to study the incidence of non-compliance with exposure classes in reinforced concrete structures. To carry out this study, surveys were carried out at several sites (districts) in Dakar (Cité Avion, Touba Ouakam, Cité Asecna, Cité Batrain, Cité Comico, Cité Assemblée and Terme Sud) in order to collect information on the formulation and implementation methods used. The comparison of the various readings carried out made it possible to deduce conclusions and to give recommendations when using standard NF EN 206-1.展开更多
Precise quantifi cation of climate-growth relationships can make a major contribution to scientifi c forest management.However,whether diff erences in the response of growth to climate at diff erent altitudes remains ...Precise quantifi cation of climate-growth relationships can make a major contribution to scientifi c forest management.However,whether diff erences in the response of growth to climate at diff erent altitudes remains unclear.To answer this,264 trees of Larix kaempferi from 88 plots,representing diff erent altitudinal ranges(1000-2100 m)and tree classes were sampled and used to develop tree-ring chronologies.Tree-ring growth(TRG)was either positively(dominant)or negatively(intermediate and suppressed)correlated with climate in diff erent tree classes at diff erent altitudes.TRG was strongly correlated with growing season at low altitudes,but was less sensitive to climate at middle altitudes.It was mainly limited by precipitation and was highly sensitive to climate at low altitudes.Climate-growth relationships at high altitudes were opposite compared to those at low altitudes.TRG of dominant trees was more sensitive to climate change compared to intermediate and suppressed trees.Climate factors(annual temperatures;moisture,the number of frost-free days)had diff erent eff ects on tree-ring growth of diff erent tree classes along altitudinal gradients.It was concluded that the increase in summer temperatures decreased water availability,resulting in a signifi cant decline in growth rates after 2005 at lower altitudes.L.kaempferi is suitable for planting in middle altitudes and dominant trees were the best sampling choice for accurately assessing climate-growth relationships.展开更多
Recently,convolutional neural network(CNN)-based visual inspec-tion has been developed to detect defects on building surfaces automatically.The CNN model demonstrates remarkable accuracy in image data analysis;however...Recently,convolutional neural network(CNN)-based visual inspec-tion has been developed to detect defects on building surfaces automatically.The CNN model demonstrates remarkable accuracy in image data analysis;however,the predicted results have uncertainty in providing accurate informa-tion to users because of the“black box”problem in the deep learning model.Therefore,this study proposes a visual explanation method to overcome the uncertainty limitation of CNN-based defect identification.The visual repre-sentative gradient-weights class activation mapping(Grad-CAM)method is adopted to provide visually explainable information.A visualizing evaluation index is proposed to quantitatively analyze visual representations;this index reflects a rough estimate of the concordance rate between the visualized heat map and intended defects.In addition,an ablation study,adopting three-branch combinations with the VGG16,is implemented to identify perfor-mance variations by visualizing predicted results.Experiments reveal that the proposed model,combined with hybrid pooling,batch normalization,and multi-attention modules,achieves the best performance with an accuracy of 97.77%,corresponding to an improvement of 2.49%compared with the baseline model.Consequently,this study demonstrates that reliable results from an automatic defect classification model can be provided to an inspector through the visual representation of the predicted results using CNN models.展开更多
调研了美国Class Central网站发布的“2023年250门全球最受欢迎在线课程榜”19门英语写作类MOOC课程。How to Write an Essay被列为250门全球最受欢迎MOOC课程之一,源于其优质的教学设计:世界顶尖大学及其丰富的在线课程经验、优秀师资...调研了美国Class Central网站发布的“2023年250门全球最受欢迎在线课程榜”19门英语写作类MOOC课程。How to Write an Essay被列为250门全球最受欢迎MOOC课程之一,源于其优质的教学设计:世界顶尖大学及其丰富的在线课程经验、优秀师资,教学目标注重培养学习者技能、终身学习理念以及抛锚式教学模式,这对于在线课程建设和教学具有借鉴意义。展开更多
Background:Due to the high heterogeneity among hepatocellular carcinoma(HCC)patients receiving transarterial chemoembolization(TACE),the prognosis of patients varies significantly.The decisionmaking on the initiation ...Background:Due to the high heterogeneity among hepatocellular carcinoma(HCC)patients receiving transarterial chemoembolization(TACE),the prognosis of patients varies significantly.The decisionmaking on the initiation and/or repetition of TACE under different liver functions is a matter of concern in clinical practice.Thus,we aimed to develop a prediction model for TACE candidates using risk stratification based on varied liver function.Methods:A total of 222 unresectable HCC patients who underwent TACE as their only treatment were included in this study.Cox proportional hazards regression was performed to select the independent risk factors and establish a predictive model for the overall survival(OS).The model was validated in patients with different Child-Pugh class and compared to previous TACE scoring systems.Results:The five independent risk factors,including alpha-fetoprotein(AFP)level,maximal tumor size,the increase of albumin-bilirubin(ALBI)grade score,tumor response,and the increase of aspartate aminotransferase(AST),were used to build a prognostic model(ASARA).In the training and validation cohorts,the OS of patients with ASARA score≤2 was significantly higher than that of patients with ASARA score>2(P<0.001,P=0.006,respectively).The ASARA model and its modified version“AS(ARA)”can effectively distinguish the OS(P<0.001,P=0.004)between patients with Child-Pugh class A and B,and the C-index was 0.687 and 0.706,respectively.For repeated TACE,the ASARA model was superior to Assessment for Retreatment with TACE(ART)and ALBI grade,maximal tumor size,AFP,and tumor response(ASAR)among Child-Pugh class A patients.For the first TACE,the performance of AS(ARA)was better than that of modified hepatoma arterial-embolization prognostic(mHAP),mHAP3,and ASA(R)models among Child-Pugh class B patients.Conclusions:The ASARA scoring system is valuable in the decision-making of TACE repetition for HCC patients,especially Child-Pugh class A patients.The modified AS(ARA)can be used to screen the ideal candidate for TACE initiation in Child-Pugh class B patients with poor liver function.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
Pneumonia is an acute lung infection that has caused many fatalitiesglobally. Radiologists often employ chest X-rays to identify pneumoniasince they are presently the most effective imaging method for this purpose.Com...Pneumonia is an acute lung infection that has caused many fatalitiesglobally. Radiologists often employ chest X-rays to identify pneumoniasince they are presently the most effective imaging method for this purpose.Computer-aided diagnosis of pneumonia using deep learning techniques iswidely used due to its effectiveness and performance. In the proposed method,the Synthetic Minority Oversampling Technique (SMOTE) approach is usedto eliminate the class imbalance in the X-ray dataset. To compensate forthe paucity of accessible data, pre-trained transfer learning is used, and anensemble Convolutional Neural Network (CNN) model is developed. Theensemble model consists of all possible combinations of the MobileNetv2,Visual Geometry Group (VGG16), and DenseNet169 models. MobileNetV2and DenseNet169 performed well in the Single classifier model, with anaccuracy of 94%, while the ensemble model (MobileNetV2+DenseNet169)achieved an accuracy of 96.9%. Using the data synchronous parallel modelin Distributed Tensorflow, the training process accelerated performance by98.6% and outperformed other conventional approaches.展开更多
Existing image captioning models usually build the relation between visual information and words to generate captions,which lack spatial infor-mation and object classes.To address the issue,we propose a novel Position...Existing image captioning models usually build the relation between visual information and words to generate captions,which lack spatial infor-mation and object classes.To address the issue,we propose a novel Position-Class Awareness Transformer(PCAT)network which can serve as a bridge between the visual features and captions by embedding spatial information and awareness of object classes.In our proposal,we construct our PCAT network by proposing a novel Grid Mapping Position Encoding(GMPE)method and refining the encoder-decoder framework.First,GMPE includes mapping the regions of objects to grids,calculating the relative distance among objects and quantization.Meanwhile,we also improve the Self-attention to adapt the GMPE.Then,we propose a Classes Semantic Quantization strategy to extract semantic information from the object classes,which is employed to facilitate embedding features and refining the encoder-decoder framework.To capture the interaction between multi-modal features,we propose Object Classes Awareness(OCA)to refine the encoder and decoder,namely OCAE and OCAD,respectively.Finally,we apply GMPE,OCAE and OCAD to form various combinations and to complete the entire PCAT.We utilize the MSCOCO dataset to evaluate the performance of our method.The results demonstrate that PCAT outperforms the other competitive methods.展开更多
BACKGROUND Patient satisfaction with facial appearance at the end of orthodontic camouflage treatment is very important, especially for skeletal malocclusion. This case report highlights the importance of the treatmen...BACKGROUND Patient satisfaction with facial appearance at the end of orthodontic camouflage treatment is very important, especially for skeletal malocclusion. This case report highlights the importance of the treatment plan for a patient initially treated with four-premolar-extraction camouflage, despite indications for orthognathic surgery.CASE SUMMARY A 23-year-old male sought treatment complaining about his unsatisfactory facial appearance. His maxillary first premolars and mandibular second premolars had been extracted, and a fixed appliance had been used to retract his anterior teeth for two years without improvement. He had a convex profile, a gummy smile, lip incompetence, inadequate maxillary incisor inclination, and almost a class I molar relationship. Cephalometric analysis showed severe skeletal class Ⅱ malocclusion(A point-nasion-B point = 11.5°) with a retrognathic mandible(sella-nasion-B point = 75.9°), a protruded maxilla(sella-nasion-A point = 87.4°), and vertical maxillary excess(upper incisor to palatal plane = 33.2 mm). The excessive lingual inclination of the maxillary incisors(upper incisor to nasion-A point line =-5.5°)was due to previous treatment attempts to compensate for the skeletal class Ⅱ malocclusion. The patient was successfully retreated with decompensating orthodontic treatment combined with orthognathic surgery. The maxillary incisors were repositioned and proclined in the alveolar bone, the overjet was increased, and a space was created for orthognathic surgery, including maxillary impaction, anterior maxillary back-setting, and bilateral sagittal split ramus osteotomy to correct his skeletal anteroposterior discrepancy. Gingival display was reduced, and lip competence was restored. In addition, the results remained stable after 2 years. The patient was satisfied with his new profile as well as with the functional malocclusion at the end of treatment.CONCLUSION This case report provides orthodontists a good example of how to treat an adult with severe skeletal class Ⅱ malocclusion with vertical maxillary excess after an unsatisfactory orthodontic camouflage treatment. Orthodontic and orthognathic treatment can significantly correct a patient’s facial appearance.展开更多
BACKGROUND Treatment for deep overbite cases can be difficult. This case report presents some techniques with improved super-elastic Ti–Ni alloy wire(ISW) for deep overbite correction.CASE SUMMARY A 21-year-old woman...BACKGROUND Treatment for deep overbite cases can be difficult. This case report presents some techniques with improved super-elastic Ti–Ni alloy wire(ISW) for deep overbite correction.CASE SUMMARY A 21-year-old woman had a chief complaint of flaring maxillary teeth. Orthodontic evaluation revealed a skeletal class Ⅱ malocclusion and a convex profile appearance. A deep overbite with palatal impingement and large overjet were also noted. Bilateral maxillary first premolars were extracted, and spaces were closed using a closed-coil spring and elastic chain. The deep overbite was corrected by applying the ISW curve and ISW intrusion arch. Intermaxillary elastics was used to adjust the intermaxillary relationship. Active treatment took approximately 3 years, and the appearance and dentition alignment noticeably improved.CONCLUSION The use of the ISW technique in a case of skeletal class Ⅱ malocclusion with deep overbite achieved a desirable result, and the patient was satisfied with the treatment outcome.展开更多
Introduction: The Six-Minute Walk Test (6MWT) is an inexpensive method to objectively evaluate physical capacity or limitation and stratify prognosis in patients with Heart Failure (HF). Since the clinical p...Introduction: The Six-Minute Walk Test (6MWT) is an inexpensive method to objectively evaluate physical capacity or limitation and stratify prognosis in patients with Heart Failure (HF). Since the clinical perception of symptoms may be adapted or compromised, regular evaluation from medical interviews often fails to determine functional classification. This study aimed to assess the correlation between New York Heart Association Functional Class (NYHA-FC) and the distance walked in the 6MWT. Methods: We conducted a cross-sectional observational study that included patients with HF with reduced ejection fraction followed up at an outpatient service of a teaching hospital, from August 2018 to April 2019. Patients in NYHA-FC I, II, or III were included. We compared NYHA-FC subjectively obtained during the consultation with the 6MWT performed after medical consultation, and the correlation between these two parameters was assessed. Results: The study included 70 patients with HF, 41 (58.6%) of whom were female. The mean age was 61.2 ± 12.7 years. The most prevalent etiologies were dilated idiopathic cardiomyopathy (35.7%) followed by ischemic cardiomyopathy (25.7%). The mean ejection fraction was 34.1% ± 9.8%. The average distance walked in the 6MWT by NYHA-FC I patients was 437.8 ± 95.8 meters, NYHA-FC II 360.1 ± 96.4, and NYHA-FC III 248.4 ± 98.3. Functional class measured by the 6MWT was different than that estimated by NYHA-FC in 34 patients (48.6%), 23 (32.9%) for a higher functional class and 11 (15.7%) for a lower one (p = 0.07). Pearson’s correlation coefficient between NYHA-FC and the 6MWT was -0.55. Conclusion: There was a moderate correlation between the subjective NYHA-FC and the 6MWT. The 6MWT revealed a different classification from NYHA-FC in almost half of the patients. Among those who presented discrepancies between methods, 6MWT reclassification towards a higher functional class was more common.展开更多
文摘One-class classification problem has become a popular problem in many fields, with a wide range of applications in anomaly detection, fault diagnosis, and face recognition. We investigate the one-class classification problem for second-order tensor data. Traditional vector-based one-class classification methods such as one-class support vector machine (OCSVM) and least squares one-class support vector machine (LSOCSVM) have limitations when tensor is used as input data, so we propose a new tensor one-class classification method, LSOCSTM, which directly uses tensor as input data. On one hand, using tensor as input data not only enables to classify tensor data, but also for vector data, classifying it after high dimensionalizing it into tensor still improves the classification accuracy and overcomes the over-fitting problem. On the other hand, different from one-class support tensor machine (OCSTM), we use squared loss instead of the original loss function so that we solve a series of linear equations instead of quadratic programming problems. Therefore, we use the distance to the hyperplane as a metric for classification, and the proposed method is more accurate and faster compared to existing methods. The experimental results show the high efficiency of the proposed method compared with several state-of-the-art methods.
基金Natural Science Foundation of Jiangsu Province, No. SBK2021021787the Major Project of the Health Commission ofJiangsu Province, No. ZD2022025and the Key Project of the Nanjing Health Commission, No. ZKX20048.
文摘BACKGROUND The relation between orthodontic treatment and temporomandibular disorders(TMDs)is under debate;the management of TMD during orthodontic treatment has always been a challenge.If TMD symptoms occur during orthodontic treatment,an immediate pause of orthodontic adjustments is recommended;the treatment can resume when the symptoms are managed and stabilized.CASE SUMMARY This case report presents a patient(26-year-old,female)with angle class I,skeletal class II and TMDs.The treatment was a hybrid of clear aligners,fixed appliances and temporary anchorage devices(TADs).After 3 mo resting and treatment on her TMD,the patient’s TMD symptom alleviated,but her anterior occlusion displayed deep overbite.Therefore,the fixed appliances with TAD were used to correct the anterior deep-bite and level maxillary and mandibular deep curves.After the levelling,the patient showed dual bite with centric relation and maximum intercuspation discrepancy on her occlusion.After careful examination of temporomandibular joints(TMJ)position,the stable bite splint and Invisible Mandibular Advancement appliance were used to reconstruct her occlusion.Eventually,the improved facial appearance and relatively stable occlusion were achieved.The 1-year follow-up records showed there was no obvious change in TMJ morphology,and her occlusion was stable.CONCLUSION TMD screening and monitoring is of great clinical importance in the TMD susceptible patients.Hybrid treatment with clear aligners and fixed appliances and TADs is an effective treatment modality for the complex cases.
文摘This study analyzed the difference between using a downward breaststroke kick and a horizontal breaststroke kick in a sample of world class elite swimmers.We compared average muscle activity of the gluteus maximus,quadriceps femoris(vastus medialis and rectus femoris),hamstring/long head of the biceps femoris,gastrocnemius medialis,rectus abdominal,and erector spinae when using the downward breaststroke kick technique.We find that when this sample of swimmers utilized the downward breaststroke kick,max speed and velocity per stroke increased,measured by 12,788 EMG samples,where the results are highly correlated to duration of the aerodynamic buoyant force in breaststroke kick technique.The increases in performance observed from measuring the world class elite swimmers is highly correlated to the duration of the kick aerodynamic buoyant force.Among this sample of elite swimmers,the longer a swimmer demonstrates a buoyant force breaststroke kick,the lower the time in a 100 breaststroke.
基金The authors would like to extend their gratitude to Universiti Teknologi PETRONAS (Malaysia)for funding this research through grant number (015LA0-037).
文摘Every application in a smart city environment like the smart grid,health monitoring, security, and surveillance generates non-stationary datastreams. Due to such nature, the statistical properties of data changes overtime, leading to class imbalance and concept drift issues. Both these issuescause model performance degradation. Most of the current work has beenfocused on developing an ensemble strategy by training a new classifier on thelatest data to resolve the issue. These techniques suffer while training the newclassifier if the data is imbalanced. Also, the class imbalance ratio may changegreatly from one input stream to another, making the problem more complex.The existing solutions proposed for addressing the combined issue of classimbalance and concept drift are lacking in understating of correlation of oneproblem with the other. This work studies the association between conceptdrift and class imbalance ratio and then demonstrates how changes in classimbalance ratio along with concept drift affect the classifier’s performance.We analyzed the effect of both the issues on minority and majority classesindividually. To do this, we conducted experiments on benchmark datasetsusing state-of-the-art classifiers especially designed for data stream classification.Precision, recall, F1 score, and geometric mean were used to measure theperformance. Our findings show that when both class imbalance and conceptdrift problems occur together the performance can decrease up to 15%. Ourresults also show that the increase in the imbalance ratio can cause a 10% to15% decrease in the precision scores of both minority and majority classes.The study findings may help in designing intelligent and adaptive solutionsthat can cope with the challenges of non-stationary data streams like conceptdrift and class imbalance.
基金supported by the National Natural Science Foundation of China (62201438,61772397,12005169)the Basic Research Program of Natural Sciences of Shaanxi Province (2021JC-23)+2 种基金Yulin Science and Technology Bureau Science and Technology Development Special Project (CXY-2020-094)Shaanxi Forestry Science and Technology Innovation Key Project (SXLK2022-02-8)the Project of Shaanxi F ederation of Social Sciences (2022HZ1759)。
文摘The development of image classification is one of the most important research topics in remote sensing. The prediction accuracy depends not only on the appropriate choice of the machine learning method but also on the quality of the training datasets. However, real-world data is not perfect and often suffers from noise. This paper gives an overview of noise filtering methods. Firstly, the types of noise and the consequences of class noise on machine learning are presented. Secondly, class noise handling methods at both the data level and the algorithm level are introduced. Then ensemble-based class noise handling methods including class noise removal, correction, and noise robust ensemble learners are presented. Finally, a summary of existing data-cleaning techniques is given.
文摘From the National Agency for the Organization and Construction of Infrastructures (ANOCI) program to the Emerging Senegal Plan (PSE), construction in Senegal has improved considerably. However, deficiencies remain in the specification of materials used mainly in hydraulic concrete. They are generally related to the specification of aggregates for alkali reactivity and the choice of exposure classes specified in NF EN 206-1. The purpose of this article is to study the incidence of non-compliance with exposure classes in reinforced concrete structures. To carry out this study, surveys were carried out at several sites (districts) in Dakar (Cité Avion, Touba Ouakam, Cité Asecna, Cité Batrain, Cité Comico, Cité Assemblée and Terme Sud) in order to collect information on the formulation and implementation methods used. The comparison of the various readings carried out made it possible to deduce conclusions and to give recommendations when using standard NF EN 206-1.
基金funded by Fundamental Research Funds of CAF (CAFYBB2022ZA00103)National Natural Science Foundation of China (General Program)(31971652)+1 种基金National Natural Science Foundation of China (32001308)Fundamental Research Funds of CAF (CAFYBB2022ZC001)
文摘Precise quantifi cation of climate-growth relationships can make a major contribution to scientifi c forest management.However,whether diff erences in the response of growth to climate at diff erent altitudes remains unclear.To answer this,264 trees of Larix kaempferi from 88 plots,representing diff erent altitudinal ranges(1000-2100 m)and tree classes were sampled and used to develop tree-ring chronologies.Tree-ring growth(TRG)was either positively(dominant)or negatively(intermediate and suppressed)correlated with climate in diff erent tree classes at diff erent altitudes.TRG was strongly correlated with growing season at low altitudes,but was less sensitive to climate at middle altitudes.It was mainly limited by precipitation and was highly sensitive to climate at low altitudes.Climate-growth relationships at high altitudes were opposite compared to those at low altitudes.TRG of dominant trees was more sensitive to climate change compared to intermediate and suppressed trees.Climate factors(annual temperatures;moisture,the number of frost-free days)had diff erent eff ects on tree-ring growth of diff erent tree classes along altitudinal gradients.It was concluded that the increase in summer temperatures decreased water availability,resulting in a signifi cant decline in growth rates after 2005 at lower altitudes.L.kaempferi is suitable for planting in middle altitudes and dominant trees were the best sampling choice for accurately assessing climate-growth relationships.
基金supported by a Korea Agency for Infrastructure Technology Advancement(KAIA)grant funded by the Ministry of Land,Infrastructure,and Transport(Grant 22CTAP-C163951-02).
文摘Recently,convolutional neural network(CNN)-based visual inspec-tion has been developed to detect defects on building surfaces automatically.The CNN model demonstrates remarkable accuracy in image data analysis;however,the predicted results have uncertainty in providing accurate informa-tion to users because of the“black box”problem in the deep learning model.Therefore,this study proposes a visual explanation method to overcome the uncertainty limitation of CNN-based defect identification.The visual repre-sentative gradient-weights class activation mapping(Grad-CAM)method is adopted to provide visually explainable information.A visualizing evaluation index is proposed to quantitatively analyze visual representations;this index reflects a rough estimate of the concordance rate between the visualized heat map and intended defects.In addition,an ablation study,adopting three-branch combinations with the VGG16,is implemented to identify perfor-mance variations by visualizing predicted results.Experiments reveal that the proposed model,combined with hybrid pooling,batch normalization,and multi-attention modules,achieves the best performance with an accuracy of 97.77%,corresponding to an improvement of 2.49%compared with the baseline model.Consequently,this study demonstrates that reliable results from an automatic defect classification model can be provided to an inspector through the visual representation of the predicted results using CNN models.
文摘调研了美国Class Central网站发布的“2023年250门全球最受欢迎在线课程榜”19门英语写作类MOOC课程。How to Write an Essay被列为250门全球最受欢迎MOOC课程之一,源于其优质的教学设计:世界顶尖大学及其丰富的在线课程经验、优秀师资,教学目标注重培养学习者技能、终身学习理念以及抛锚式教学模式,这对于在线课程建设和教学具有借鉴意义。
基金This study was supported by a grant from Tianjin Key Medical Discipline(Specialty)Construction Project.
文摘Background:Due to the high heterogeneity among hepatocellular carcinoma(HCC)patients receiving transarterial chemoembolization(TACE),the prognosis of patients varies significantly.The decisionmaking on the initiation and/or repetition of TACE under different liver functions is a matter of concern in clinical practice.Thus,we aimed to develop a prediction model for TACE candidates using risk stratification based on varied liver function.Methods:A total of 222 unresectable HCC patients who underwent TACE as their only treatment were included in this study.Cox proportional hazards regression was performed to select the independent risk factors and establish a predictive model for the overall survival(OS).The model was validated in patients with different Child-Pugh class and compared to previous TACE scoring systems.Results:The five independent risk factors,including alpha-fetoprotein(AFP)level,maximal tumor size,the increase of albumin-bilirubin(ALBI)grade score,tumor response,and the increase of aspartate aminotransferase(AST),were used to build a prognostic model(ASARA).In the training and validation cohorts,the OS of patients with ASARA score≤2 was significantly higher than that of patients with ASARA score>2(P<0.001,P=0.006,respectively).The ASARA model and its modified version“AS(ARA)”can effectively distinguish the OS(P<0.001,P=0.004)between patients with Child-Pugh class A and B,and the C-index was 0.687 and 0.706,respectively.For repeated TACE,the ASARA model was superior to Assessment for Retreatment with TACE(ART)and ALBI grade,maximal tumor size,AFP,and tumor response(ASAR)among Child-Pugh class A patients.For the first TACE,the performance of AS(ARA)was better than that of modified hepatoma arterial-embolization prognostic(mHAP),mHAP3,and ASA(R)models among Child-Pugh class B patients.Conclusions:The ASARA scoring system is valuable in the decision-making of TACE repetition for HCC patients,especially Child-Pugh class A patients.The modified AS(ARA)can be used to screen the ideal candidate for TACE initiation in Child-Pugh class B patients with poor liver function.
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.
文摘Pneumonia is an acute lung infection that has caused many fatalitiesglobally. Radiologists often employ chest X-rays to identify pneumoniasince they are presently the most effective imaging method for this purpose.Computer-aided diagnosis of pneumonia using deep learning techniques iswidely used due to its effectiveness and performance. In the proposed method,the Synthetic Minority Oversampling Technique (SMOTE) approach is usedto eliminate the class imbalance in the X-ray dataset. To compensate forthe paucity of accessible data, pre-trained transfer learning is used, and anensemble Convolutional Neural Network (CNN) model is developed. Theensemble model consists of all possible combinations of the MobileNetv2,Visual Geometry Group (VGG16), and DenseNet169 models. MobileNetV2and DenseNet169 performed well in the Single classifier model, with anaccuracy of 94%, while the ensemble model (MobileNetV2+DenseNet169)achieved an accuracy of 96.9%. Using the data synchronous parallel modelin Distributed Tensorflow, the training process accelerated performance by98.6% and outperformed other conventional approaches.
基金supported by the National Key Research and Development Program of China[No.2021YFB2206200].
文摘Existing image captioning models usually build the relation between visual information and words to generate captions,which lack spatial infor-mation and object classes.To address the issue,we propose a novel Position-Class Awareness Transformer(PCAT)network which can serve as a bridge between the visual features and captions by embedding spatial information and awareness of object classes.In our proposal,we construct our PCAT network by proposing a novel Grid Mapping Position Encoding(GMPE)method and refining the encoder-decoder framework.First,GMPE includes mapping the regions of objects to grids,calculating the relative distance among objects and quantization.Meanwhile,we also improve the Self-attention to adapt the GMPE.Then,we propose a Classes Semantic Quantization strategy to extract semantic information from the object classes,which is employed to facilitate embedding features and refining the encoder-decoder framework.To capture the interaction between multi-modal features,we propose Object Classes Awareness(OCA)to refine the encoder and decoder,namely OCAE and OCAD,respectively.Finally,we apply GMPE,OCAE and OCAD to form various combinations and to complete the entire PCAT.We utilize the MSCOCO dataset to evaluate the performance of our method.The results demonstrate that PCAT outperforms the other competitive methods.
文摘BACKGROUND Patient satisfaction with facial appearance at the end of orthodontic camouflage treatment is very important, especially for skeletal malocclusion. This case report highlights the importance of the treatment plan for a patient initially treated with four-premolar-extraction camouflage, despite indications for orthognathic surgery.CASE SUMMARY A 23-year-old male sought treatment complaining about his unsatisfactory facial appearance. His maxillary first premolars and mandibular second premolars had been extracted, and a fixed appliance had been used to retract his anterior teeth for two years without improvement. He had a convex profile, a gummy smile, lip incompetence, inadequate maxillary incisor inclination, and almost a class I molar relationship. Cephalometric analysis showed severe skeletal class Ⅱ malocclusion(A point-nasion-B point = 11.5°) with a retrognathic mandible(sella-nasion-B point = 75.9°), a protruded maxilla(sella-nasion-A point = 87.4°), and vertical maxillary excess(upper incisor to palatal plane = 33.2 mm). The excessive lingual inclination of the maxillary incisors(upper incisor to nasion-A point line =-5.5°)was due to previous treatment attempts to compensate for the skeletal class Ⅱ malocclusion. The patient was successfully retreated with decompensating orthodontic treatment combined with orthognathic surgery. The maxillary incisors were repositioned and proclined in the alveolar bone, the overjet was increased, and a space was created for orthognathic surgery, including maxillary impaction, anterior maxillary back-setting, and bilateral sagittal split ramus osteotomy to correct his skeletal anteroposterior discrepancy. Gingival display was reduced, and lip competence was restored. In addition, the results remained stable after 2 years. The patient was satisfied with his new profile as well as with the functional malocclusion at the end of treatment.CONCLUSION This case report provides orthodontists a good example of how to treat an adult with severe skeletal class Ⅱ malocclusion with vertical maxillary excess after an unsatisfactory orthodontic camouflage treatment. Orthodontic and orthognathic treatment can significantly correct a patient’s facial appearance.
文摘BACKGROUND Treatment for deep overbite cases can be difficult. This case report presents some techniques with improved super-elastic Ti–Ni alloy wire(ISW) for deep overbite correction.CASE SUMMARY A 21-year-old woman had a chief complaint of flaring maxillary teeth. Orthodontic evaluation revealed a skeletal class Ⅱ malocclusion and a convex profile appearance. A deep overbite with palatal impingement and large overjet were also noted. Bilateral maxillary first premolars were extracted, and spaces were closed using a closed-coil spring and elastic chain. The deep overbite was corrected by applying the ISW curve and ISW intrusion arch. Intermaxillary elastics was used to adjust the intermaxillary relationship. Active treatment took approximately 3 years, and the appearance and dentition alignment noticeably improved.CONCLUSION The use of the ISW technique in a case of skeletal class Ⅱ malocclusion with deep overbite achieved a desirable result, and the patient was satisfied with the treatment outcome.
文摘Introduction: The Six-Minute Walk Test (6MWT) is an inexpensive method to objectively evaluate physical capacity or limitation and stratify prognosis in patients with Heart Failure (HF). Since the clinical perception of symptoms may be adapted or compromised, regular evaluation from medical interviews often fails to determine functional classification. This study aimed to assess the correlation between New York Heart Association Functional Class (NYHA-FC) and the distance walked in the 6MWT. Methods: We conducted a cross-sectional observational study that included patients with HF with reduced ejection fraction followed up at an outpatient service of a teaching hospital, from August 2018 to April 2019. Patients in NYHA-FC I, II, or III were included. We compared NYHA-FC subjectively obtained during the consultation with the 6MWT performed after medical consultation, and the correlation between these two parameters was assessed. Results: The study included 70 patients with HF, 41 (58.6%) of whom were female. The mean age was 61.2 ± 12.7 years. The most prevalent etiologies were dilated idiopathic cardiomyopathy (35.7%) followed by ischemic cardiomyopathy (25.7%). The mean ejection fraction was 34.1% ± 9.8%. The average distance walked in the 6MWT by NYHA-FC I patients was 437.8 ± 95.8 meters, NYHA-FC II 360.1 ± 96.4, and NYHA-FC III 248.4 ± 98.3. Functional class measured by the 6MWT was different than that estimated by NYHA-FC in 34 patients (48.6%), 23 (32.9%) for a higher functional class and 11 (15.7%) for a lower one (p = 0.07). Pearson’s correlation coefficient between NYHA-FC and the 6MWT was -0.55. Conclusion: There was a moderate correlation between the subjective NYHA-FC and the 6MWT. The 6MWT revealed a different classification from NYHA-FC in almost half of the patients. Among those who presented discrepancies between methods, 6MWT reclassification towards a higher functional class was more common.