Multi-level Classification Research Articles

A Binary Multi Class and Multi Level Classification with Dual Priority Labelling Model for COVID-19 and Other Thorax Disease Detection

A Binary Multi Class and Multi Level Classification with Dual Priority Labelling Model for COVID-19 and Other Thorax Disease Detection

Diagnosis of heart failure from imbalance datasets using multi-level classification

Background:The incidence of heart failure is continuing to rise, and the mortality rate is high. Chest X-ray (CXR) has irreplaceable advantages in diagnosing heart failure, such as fast, low risk, and cheap. However, excessive CXR images place a huge burden on physicians and create data imbalance problems. Traditional methods, such as random under-sampling, are used to solve the problems. However, the under-sampling method can destroy the integrity of the data distribution. So, it is necessary to have a method that can address data imbalance problems and assist overburdened healthcare systems. Objective:This study establishes an automatic heart failure diagnosis system based on deep learning from imbalance datasets. Methods:To address the data imbalance problem based on the publicly available CheXpert dataset, this study proposes a method combining under-sampling and instance selection to ensure the integrity of the data distribution. To help physicians better treat heart failure, this study proposes an end-to-end multi-level classification method to diagnose the specific causes of heart failure. Results:On the testing set, our method improves the average accuracy by 3.78% compared to the traditional random under-sampling method, and the accuracy of our end-to-end multi-class classification experiment is 84.44%. Conclusions:The heart failure automated diagnostic system is more efficient and accurate in diagnosing heart failure compared to state-of-the-art methods.

New insight into fate and transport of organic compounds from pollution sources to aquatic environment using non-targeted screening: A wastewater treatment plant case study

A variety of chemicals discharged into the aquatic environment by the wastewater treatment plant (WWTP), which is a potential source of hazard to the ecological environment and human health. This study established a novel analytical method for all compounds using non-targeted screening to comprehensively explore the fate and transport of organic compounds from WWTP to aquatic environment. 3967 and 3636 features were detected in WWTP samples and river samples, respectively. Multi-level classification was applied to all identified compounds, and results showed that aliphatics were dominant in both abundance and response, accounting for an average of 35.49 % and 74.10 %, respectively. A total of 88 Emerging Contaminants (ECs), including 22 endocrine disrupting chemicals (EDCs), 12 pharmaceuticals and personal care products (PPCPs), 12 pesticides, 10 volatile organic compounds (VOCs), 5 persistent organic pollutants (POPs) and 27 chemicals with other uses, were identified from all compounds, and their traceability analysis was performed. Furthermore, the contribution rate of organic compounds from WWTP effluent to river was calculated to be 33.60 % by the analysis of source-sink relationship. An in-depth and comprehensive exploration of the fate and transport of all organic compounds will help to provide guidelines for the treatment technologies and achieve the traceability of pollutants.

Multi-Stage Classification-Based Deep Learning for Gleason System Grading Using Histopathological Images.

In this work, we introduced an automated diagnostic system for Gleason system grading and grade groups (GG) classification using whole slide images (WSIs) of digitized prostate biopsy specimens (PBSs). Our system first classifies the Gleason pattern (GP) from PBSs and then identifies the Gleason score (GS) and GG. We developed a comprehensive DL-based approach to develop a grading pipeline system for the digitized PBSs and consider GP as a classification problem (not segmentation) compared to current research studies (deals with as a segmentation problem). A multilevel binary classification was implemented to enhance the segmentation accuracy for GP. Also, we created three levels of analysis (pyramidal levels) to extract different types of features. Each level has four shallow binary CNN to classify five GP labels. A majority fusion is applied for each pixel that has a total of 39 labeled images to create the final output for GP. The proposed framework is trained, validated, and tested on 3080 WSIs of PBS. The overall diagnostic accuracy for each CNN is evaluated using several metrics: precision (PR), recall (RE), and accuracy, which are documented by the confusion matrices.The results proved our system's potential for classifying all five GP and, thus, GG. The overall accuracy for the GG is evaluated using two metrics, PR and RE. The grade GG results are between 50% to 92% for RE and 50% to 92% for PR. Also, a comparison between our CNN architecture and the standard CNN (ResNet50) highlights our system's advantage. Finally, our deep-learning system achieved an agreement with the consensus grade groups.

Multilevel hybrid accurate handcrafted model for myocardial infarction classification using ECG signals.

Myocardial infarction (MI) is detected using electrocardiography (ECG) signals. Machine learning (ML) models have been used for automated MI detection on ECG signals. Deep learning models generally yield high classification performance but are computationally intensive. We have developed a novel multilevel hybrid feature extraction-based classification model with low time complexity for MI classification. The study dataset comprising 12-lead ECGs belonging to one healthy and 10 MI classes were downloaded from a public ECG signal databank. The model architecture comprised multilevel hybrid feature extraction, iterative feature selection, classification, and iterative majority voting (IMV). In the hybrid handcrafted feature (HHF) generation phase, both textural and statistical feature extraction functions were used to extract features from ECG beats but only at a low level. A new pooling-based multilevel decomposition model was presented to enable them to create features at a high level. This model used average and maximum pooling to create decomposed signals. Using these pooling functions, an unbalanced tree was obtained. Therefore, this model was named multilevel unbalanced pooling tree transformation (MUPTT). On the feature extraction side, two extractors (functions) were used to generate both statistical and textural features. To generate statistical features, 20 commonly used moments were used. A new, improved symmetric binary pattern function was proposed to generate textural features. Both feature extractors were applied to the original MI signal and the decomposed signals generated by the MUPTT. The most valuable features from among the extracted feature vectors were selected using iterative neighborhood component analysis (INCA). In the classification phase, a one-dimensional nearest neighbor classifier with ten-fold cross-validation was used to obtain lead-wise results. The computed lead-wise results derived from all 12 leads of the same beat were input to the IMV algorithm to generate ten voted results. The most representative was chosen using a greedy technique to calculate the overall classification performance of the model. The HHF-MUPTT-based ECG beat classification model attained excellent performance, with the best lead-wise accuracy of 99.85% observed in Lead III and 99.94% classification accuracy using the IMV algorithm. The results confirmed the high MI classification ability of the presented computationally lightweight HHF-MUPTT-based model.

Coastline extraction based on multi-scale segmentation and multi-level inheritance classification

Detailed management of the coastline is critical to the development of coastal states. However, the current classification of the coastline is relatively weak. This study proposed an automatic method to detect coastlines with category attributes based on multi-scale segmentation and multi-level inheritance classification. Fully integrating the advantages of multi-scale segmentation and multi-level classification, it solved the problems that traditional methods could not solve, such as extracting coastlines with categorical attributes, cultivation ponds that are easily affected by tidal flats, and complex coastal terrain. The Chinese GF-2 satellite images are used to extract various types of coastlines in Jiaozhou Bay and its surrounding areas such as the harbor-wharf coastline, silt coastline, pond coastline, rocky coastline, and sandy coastline. Compared with the human interpretation, it is found that the coastline extracted by our proposed method is different by 10.104 km in the harbor-wharf coastline, 0.099 km in the silt coastline, 2.677 km in the pond coastline, 8.831 km in the rocky coastline, and 0.218 km in the sandy coastline. Furthermore, compared to the object-based region growing integrating edge detection (OBRGIE) method, it is increased by 13.52%, 2.16%, 14.48%, 52.57%, and 22.97%, respectively. The results show that our proposed method is algorithmically more reasonable, accurate, and powerful. It can provide data support for refined coastline management.

A Framework for Auditing Multilevel Models using Explainability Methods

Multilevel models (MLMs) are increasingly deployed in industry across different functions. Applications usually result in binary classification within groups or hierarchies based on a set of input features. For transparent and ethical applications of such models, sound audit frameworks need to be developed. In this paper, an audit framework for technical assessment of regression MLMs is proposed. The focus is on three aspects: model, discrimination, and transparency & explainability. These aspects are subsequently divided into sub-aspects. Contributors, such as inter MLM-group fairness, feature contribution order, and aggregated feature contribution, are identified for each of these sub-aspects. To measure the performance of the contributors, the framework proposes a shortlist of KPIs, among others, intergroup individual fairness (DiffInd_MLM) across MLM-groups, probability unexplained (PUX) and percentage of incorrect feature signs (POIFS). A traffic light risk assessment method is furthermore coupled to these KPIs. For assessing transparency & explainability, different explainability methods (SHAP and LIME) are used, which are compared with a model intrinsic method using quantitative methods and machine learning modelling. Using an open-source dataset, a model is trained and tested and the KPIs are computed. It is demonstrated that popular explainability methods, such as SHAP and LIME, underperform in accuracy when interpreting these models. They fail to predict the order of feature importance, the magnitudes, and occasionally even the nature of the feature contribution (negative versus positive contribution on the outcome). For other contributors, such as group fairness and their associated KPIs, similar analysis and calculations have been performed with the aim of adding profundity to the proposed audit framework. The framework is expected to assist regulatory bodies in performing conformity assessments of AI systems using multilevel binomial classification models at businesses. It will also benefit providers, users, and assessment bodies, as defined in the European Commission’s proposed Regulation on Artificial Intelligence, when deploying AI-systems such as MLMs, to be future-proof and aligned with the regulation.

A Framework for Auditing Multilevel Models using Explainability Methods

Multilevel models (MLMs) are increasingly deployed in industry across different functions. Applications usually result in binary classification within groups or hierarchies based on a set of input features. For transparent and ethical applications of such models, sound audit frameworks need to be developed. In this paper, an audit framework for technical assessment of regression MLMs is proposed. The focus is on three aspects: model, discrimination, and transparency & explainability. These aspects are subsequently divided into sub-aspects. Contributors, such as inter MLM-group fairness, feature contribution order, and aggregated feature contribution, are identified for each of these sub-aspects. To measure the performance of the contributors, the framework proposes a shortlist of KPIs, among others, intergroup individual fairness (DiffInd_MLM) across MLM-groups, probability unexplained (PUX) and percentage of incorrect feature signs (POIFS). A traffic light risk assessment method is furthermore coupled to these KPIs. For assessing transparency & explainability, different explainability methods (SHAP and LIME) are used, which are compared with a model intrinsic method using quantitative methods and machine learning modelling.
 Using an open-source dataset, a model is trained and tested and the KPIs are computed. It is demonstrated that popular explainability methods, such as SHAP and LIME, underperform in accuracy when interpreting these models. They fail to predict the order of feature importance, the magnitudes, and occasionally even the nature of the feature contribution (negative versus positive contribution on the outcome). For other contributors, such as group fairness and their associated KPIs, similar analysis and calculations have been performed with the aim of adding profundity to the proposed audit framework. The framework is expected to assist regulatory bodies in performing conformity assessments of AI systems using multilevel binomial classification models at businesses. It will also benefit providers, users, and assessment bodies, as defined in the European Commission’s proposed Regulation on Artificial Intelligence, when deploying AI-systems such as MLMs, to be future-proof and aligned with the regulation.

The multi-level classification network (MCN) with modified residual U-Net for ischemic stroke lesions segmentation from ATLAS

Ischemic and hemorrhagic strokes are two major types of internal brain injury. 3D brain MRI is suggested by neurologists to examine the brain. Manual examination of brain MRI is very sensitive and time-consuming task. However, automatic diagnosis can assist doctors in this regard. Anatomical Tracings of Lesions After Stroke (ATLAS) is publicly available dataset for research experiments. One of the major issues in medical imaging is class imbalance. Similarly, pixel representation of stroke lesion is less than 1% in ATLAS. Second major challenge in this dataset is inter-class similarity. A multi-level classification network (MCN) is proposed for segmentation of ischemic stroke lesions. MCN consists of three cascaded discrete networks. The first network designed to reduce the slice level class imbalance, where a classifier model is trained to extract the slices of stroke lesions from a whole brain MRI volume. The interclass similarity cause to produce false positives in segmented output. Therefore, all extracted stroke slices were divided into overlapping patches (64 × 64) and carried to the second network. The task associated with second network is to classify the patches comprises of stroke lesion. The third network is a 2D modified residual U-Net that segments out the stroke lesions from the patches extracted by the second network. MCN achieved 0.754 mean dice score on test dataset which is higher than the other state-of-the-art methods on the same dataset.

Nested segmentation and multi-level classification of diabetic foot ulcer based on mask R-CNN

Nested segmentation and multi-level classification of diabetic foot ulcer based on mask R-CNN

Multi-Level Circulation Pattern Classification Based on the Transfer Learning CNN Network

Deep learning artificial intelligence technology, which has the advantages of nonlinear mapping ability, massive information extraction ability, spatial-temporal modeling ability, and so on, provides new ideas and methods for further improving the accuracy of weather and climate extreme event prediction. A transfer learning CNN (Convolutional Neural Networks) classification model is established to classify the circulation patterns, along with the newly reconstructed dataset of regional persistent historical heavy rain events, daily rainfall data of 2474 observational stations, and the NCEP/NCAR global reanalysis data of daily geopotential height field in 1981–2018. Different from previous classifications, usually with one level variable, here, in addition to 500 hPa heights, 200 hPa zonal winds and 850 hPa meridional winds over the key areas are also considered in the model. The results show that the multi-level circulation pattern classification based on the transfer learning CNN network has a higher accuracy in the independent test than the single-level model, with the accuracy reaching 92.5% (while only 85% for the single-level model). The spatial correlation coefficient of precipitation between each typical mode and related patterns obtained by the multi-level transfer learning CNN classification is greater than that obtained by the single-level transfer learning CNN, and the variance of 500 hPa heights between each typical mode and the associated patterns is also greater than that obtained by the single-level transfer learning CNN. These results show that the performance of the classification by the multi-level transfer learning CNN model is better than that by the single-level transfer learning CNN. The study is helpful to develop circulation classifications related to large-scale weather or climate disaster events and then to provide a physical basis for further improving the forecast effect and extending the valid time of the forecast through combining the numerical model products.

Label Metric for Multi-Class Multi-Target Tracking under Hierarchical Multilevel Classification.

Aiming at multiple quantities and types of targets, multi-class multi-target tracking usually faces not only cardinality errors, but also mis-classification problems. Considering its performance evaluation, the traditional optimal subpattern assignment (OSPA) method tends to calculate a separate metric for each class of targets, or introduce other indexes such as the classification error rate, which decreases the value of OSPA as a comprehensive single metric. This paper proposed a hierarchical multi-level class label for multi-class multi-target tracking under hierarchical multilevel classification, which can synthetically measure the state errors, cardinality error, and mis-classification. The hierarchical multi-level class label is introduced as an attached label to finite sets based on the hierarchical tree-structured categorization. A Wasserstein distance type metric then can be defined among the distribution represented by any two labels. The proposed label metric is a mathematic metric, and its advantages are illustrated by examples in several cases.

Системный анализ факторов, влияющих на возникновение лесных пожаров в Северо-Западном федеральном округе

A systematic analysis of the factors causing forest fires (anthropogenic and weather conditions, types 
 of forests) has been carried out. Multi-level classifications are defined for each set of factors with the inclusion 
 of terms that allow the decision makers to work with their natural language. An integral forest fire risk assessment is proposed, which is described by a fuzzy model containing a term-set. The structure of the system of criteria by which it is possible to determine the risk of fire has been developed. The proposed integral assessment can serve as an addition to the existing classical scale for assessing fire hazard in forests. It includes the parameters characterizing weather conditions and is used in a large number of countries by supervisory authorities. To differentiate the assessment of fire safety in Russia, the parameters of forestry enterprises in the study area are also taken into account. Based on the proposed matrix structures, a primary database has been developed, which in the future will be filled automatically from open sources of state information systems. The data in the systems are generated at different intervals, most of them daily, which allows for high-quality monitoring and control over the changes in the current situation in the appointed region. Based on the fire hazard scale, it will be possible to obtain an optimal assessment by analyzing a larger number of factors than in the classical assessment. For the subsequent implementation of the decision support method, matrices have been developed using the open data from state information systems. Applying the developed and completed matrices as a single data set is possible with the help of a decision support module, which is planned to be further implemented using the Python language.

Analytical review of methods for automatic detection of user engagement in virtual communication

Introduction: The solution of the task of the recognition and assessment of user engagement in the acts of human-machine interaction or telecommunication, achieved through the use of automatic means, is highly important in computer recognition of human psycho-emotional states. This is necessary for e-learning, business and entertainment applications design. Purpose: To conduct a comparative analysis of the current information support in the field of automatic recognition and assessment of user involvement in human-machine interaction or virtual communication, as well as to establish a methodology for building a data body based on the idea of multimodal communication. Results: The conducted analysis of research papers has shown that in most existing databases there is a substantial lack of data for natural online communication. Moreover, not all databases contain different modalities in “human-machine-human” communication system. Text and audio modalities turn out to be important for a multilevel engagement classification task, aimed at the determination of engagement intensity. It is also promising to take into account “body language” features, such as facial expressions, movements of the body and the head, gestures. For the correct assessment of involvement, an engagement database must contain meta-data on the psycho-emotional states of communicants. Neural network-based approaches to the automatic detection of user engagement show the best performance. Practical relevance: Based on the obtained analytical conclusions, the authors of the paper are going to elaborate an original software system for automatic recognition of user engagement, and to collect a data set for machine learning purposes. The presented review formulates basic requirements for such systems and contributes to the solution of the problem of automatic recognition of psycho-emotional states. Discussion: The survey leads to the conclusion that the notion of engagement as understood in studies on automatic emotion recognition differs from that used in psychology. User (or communicant) engagement in terms of info- and communicative sphere implies the manifestation of a person's mental activity level (emotional, cognitive, and behavioral components) changing dynamically while interacting with another person or computer system.

Multi-view and Multi-level network for fault diagnosis accommodating feature transferability

Various deep transfer learning solutions have been developed for machine fault diagnosis. The existing solutions mainly focus on domain adaptation by minimizing the data distribution discrepancy with certain metric, which emphasize the common features embedded in the data cross domains and neglect the unique features toward health condition classification in one specific domain. In these solutions, all the data for training have been forced to align in a common feature space and all the features for domain adaptation have been treated equally. However, there might exist domain specific features which are not appropriate for transfer but may contain essential information for classification in specific domain. In addition, due to the difficulty of collecting machine fault data, the number of machine fault samples is usually quite small or even zero. The traditional deep network structures and the training strategy are not the optimal choice in this occasion. To address these problems, a novel multi-view and multi-level network (MMNet) for fault diagnosis is developed. In MMNet, two network channels have been respectively constructed for cross domain common feature and domain specific feature learning to provide multi-view features. This architecture could implicitly differentiate the common features cross domains and the specific features only in one domain. In the channel of domain specific feature, a domain classifier and fault classifier are combined to learn the domain specific features. Multiple kernel maximum mean discrepancy (MK-MMD) is imposed on multiple layers of the common feature channel to implement domain adaptation and extract cross domain common features. The domain classification and fault classification together form a multi-level classification scheme. A classic few shot learning architecture with two modules respectively for feature extraction and relation computation is adopted as the backbone network. The relation score based classification mechanism enables zero shot fault classification in the target domain. Episode based few shot training strategy is employed to enhance the performance of MMNet with few labeled training data. Extensive experiments have demonstrated the state-of-the-art performance of MMNet on the involved transfer tasks.

Real time railway track crack analysis using multi-level classification

Rail transportation system remains one of the most cost effective and suited means of passenger and goods transportation for both long distance and suburban travel. Hence it is a critical task to ensure regular railway maintenance. This work explores the scope of detecting cracks and also pedestrians in the track by making use of deep learning and image recognition system. This task becomes very challenging as the cracks in the track are varying and the pedestrian can take wide range of poses. This system make use of a conventional Convolutional Neural Network (CNN) along with Support Vector Machine for recognition of crack defective from the dataset. SVM is the binary classifier used, while CNN was used as an automated feature extractor. The dataset was used to train and test the algorithm that was implemented in the proposed model. CNN’s receptive field aids in extracting the most recognizable aspects from these images automatically. The experimental findings show that the suggested framework is successful, with a recognition accuracy of 78% throughout the dataset and a template based lane detection method is used. Lane tracking is done with edge detection and the position of a region of interest in vanishing point in tack image is used for identifying the land departure. This detector on real time video with the region of interest selection and achieve in 23 fps.

Automated Extraction of Forest Burn Severity Based on Light and Small UAV Visible Remote Sensing Images

Identification of forest burn severity is essential for fire assessments and a necessary procedure in modern forest management. Due to the low efficiency and labor intensity of the current post-fire field survey in China’s Forestry Standards, the limitation of temporal resolution of satellite imagery, and poor objectivity of manual interpretations, a new method for automatic extraction of forest burn severity based on small visible unmanned aerial vehicle (UAV) images is proposed. Taking the forest fires which occurred in Anning city of Yunnan Province in 2019 as the study objects, post-fire imagery was obtained by a small, multi-rotor near-ground UAV. Some image recognition indices reflecting the variations in chlorophyll loss effects in different damaged forests were developed with spectral characteristics customized in A and C, and the texture features such as the mean, standard deviation, homogeneity, and shape index of the length–width ratio. An object-oriented method is used to determine the optimal segmentation scale for forest burn severity and a multilevel rule classification and extraction model is established to achieve the automatic identification and mapping. The results show that the method mentioned above can recognize different types of forest burn severity: unburned, damaged, dead, and burnt. The overall accuracy is 87.76%, and the Kappa coefficient is 0.8402, which implies that the small visible UAV can be used as a substitution for the current forest burn severity survey standards. This research is of great practical significance for improving the efficiency and precision of forest fire investigation, expanding applications of small UAVs in forestry, and developing an alternative for forest fire loss assessments in the forestry industry.

EEG/fNIRS Based Workload Classification Using Functional Brain Connectivity and Machine Learning.

There is high demand for techniques to estimate human mental workload during some activities for productivity enhancement or accident prevention. Most studies focus on a single physiological sensing modality and use univariate methods to analyse multi-channel electroencephalography (EEG) data. This paper proposes a new framework that relies on the features of hybrid EEG–functional near-infrared spectroscopy (EEG–fNIRS), supported by machine-learning features to deal with multi-level mental workload classification. Furthermore, instead of the well-used univariate power spectral density (PSD) for EEG recording, we propose using bivariate functional brain connectivity (FBC) features in the time and frequency domains of three bands: delta (0.5–4 Hz), theta (4–7 Hz) and alpha (8–15 Hz). With the assistance of the fNIRS oxyhemoglobin and deoxyhemoglobin (HbO and HbR) indicators, the FBC technique significantly improved classification performance at a 77% accuracy for 0-back vs. 2-back and 83% for 0-back vs. 3-back using a public dataset. Moreover, topographic and heat-map visualisation indicated that the distinguishing regions for EEG and fNIRS showed a difference among the 0-back, 2-back and 3-back test results. It was determined that the best region to assist the discrimination of the mental workload for EEG and fNIRS is different. Specifically, the posterior area performed the best for the posterior midline occipital (POz) EEG in the alpha band and fNIRS had superiority in the right frontal region (AF8).

Automatic interpretation and clinical evaluation for fundus fluorescein angiography images of diabetic retinopathy patients by deep learning

Background/aimsFundus fluorescein angiography (FFA) is an important technique to evaluate diabetic retinopathy (DR) and other retinal diseases. The interpretation of FFA images is complex and time-consuming, and the ability of...

Identification of Plant Diseases Using Multi-Level Classification Deep Model

As plants are exposed to the various environmental elements, they are highly prone to many diseases which affect the crop yield and result in low productivity. Due to the lack of proper care and regular checking for any diseases in plants, severe consequences may be seen in a long-term basis on plants and environments. Agricultural productivity is one of the important factors on which the economy highly depends. Plant pathologists require a reliable and effective method to diagnose the disease effectively. Several physical methods and techniques have been applied to better predict and classify the plant disease. However, we need an automated method to identify and produce as accurate result as possible with minimum time. Previously, many machine learning models were developed, producing limited accuracy. But, using deep learning, improved performance can be achieved for classification of plant diseases. The authors propose the multi-level classification model for plant diseases detection. The accuracy achieved by the proposed model is 96.70%, which is higher than the other models.