Search

Uniform Resource Locator (URL) is a unique identifier composed of protocol and domain name used to locate and retrieve a resource on the Internet. Like any Internet service, URLs (also called websites) are vulnerable to compromise by attackers to develop Malicious URLs that can exploit/devastate the user's information and resources. Malicious URLs are usually designed with the intention of promoting cyber-attacks such as spam, phishing, malware, and defacement. These websites usually require action on the user's side and can reach users across emails, text messages, pop-ups, or devious advertisements. They have a potential impact that can reach, in some cases, to compromise the machine or network of the user, especially those arriving by email. Therefore, developing systems to detect malicious URLs is of great interest nowadays. This paper proposes a high-performance machine learning-based detection system to identify Malicious URLs. The proposed system provides two layers of detection. Firstly, we identify the URLs as either benign or malware using a binary classifier. Secondly, we classify the URL classes based on their feature into five classes: benign, spam, phishing, malware, and defacement. Specifically, we report on four ensemble learning approaches, viz. the ensemble of bagging trees (En_Bag) approach, the ensemble of k-nearest neighbor (En_kNN) approach, and the ensemble of boosted decision trees (En_Bos) approach, and the ensemble of subspace discriminator (En_Dsc) approach. The developed approaches have been evaluated on an inclusive and contemporary dataset for uniform resource locators (ISCX-URL2016). ISCX-URL2016 provides a lightweight dataset for detecting and categorizing malicious URLs according to their attack type and lexical analysis. Conventional machine learning evaluation measurements are used to evaluate the detection accuracy, precision, recall, F Score, and detection time. Our experiential assessment indicates that the ensemble of bagging trees (En_Bag) approach provides better performance rates than other ensemble methods. Alternatively, the ensemble of the k-nearest neighbor (En_kNN) approach provides the highest inference speed. We also contrast our En_Bag model with state-of-the-art solutions and show its superiority in binary classification and multi-classification with accuracy rates of 99.3% and 97.92%, respectively. [ABSTRACT FROM AUTHOR]

Deep Neural Networks (DNNs) are widely regarded as the most effective learning tool for dealing with large datasets, and they have been successfully used in thousands of applications in a variety of fields. Based on these large datasets, they are trained to learn the relationships between various variables. The adaptive moment estimation (Adam) algorithm, a highly efficient adaptive optimization algorithm, is widely used as a learning algorithm in various fields for training DNN models. However, it needs to improve its generalization performance, especially when training with large-scale datasets. Therefore, in this paper, we propose HN Adam, a modified version of the Adam Algorithm, to improve its accuracy and convergence speed. The HN_Adam algorithm is modified by automatically adjusting the step size of the parameter updates over the training epochs. This automatic adjustment is based on the norm value of the parameter update formula according to the gradient values obtained during the training epochs. Furthermore, a hybrid mechanism was created by combining the standard Adam algorithm and the AMSGrad algorithm. As a result of these changes, the HN_Adam algorithm, like the stochastic gradient descent (SGD) algorithm, has good generalization performance and achieves fast convergence like other adaptive algorithms. To test the proposed HN_Adam algorithm performance, it is evaluated to train a deep convolutional neural network (CNN) model that classifies images using two different standard datasets: MNIST and CIFAR-10. The algorithm results are compared to the basic Adam algorithm and the SGD algorithm, in addition to other five recent SGD adaptive algorithms. In most comparisons, the HN Adam algorithm outperforms the compared algorithms in terms of accuracy and convergence speed. AdaBelief is the most competitive of the compared algorithms. In terms of testing accuracy and convergence speed (represented by the consumed training time), the HN-Adam algorithm outperforms the AdaBelief algorithm by an improvement of 1.0% and 0.29% for the MNIST dataset, and 0.93% and 1.68% for the CIFAR-10 dataset, respectively. [ABSTRACT FROM AUTHOR]

The operation of large utility companies such as Consolidated Edison Company of New York, Inc. (Con Edison) typically rely on large quantities of regulation documents from external institutions which inform the company of upcoming or ongoing policy changes or new requirements the company might need to comply with if deemed applicable. As a concrete example, if a recent regulatory publication mentions that the timeframe for the Company to respond to a reported system emergency in its service territory changes from within X time to within Y time—then the affected operating groups will be notified, and internal Company operating procedures may need to be reviewed and updated accordingly to comply with the new regulatory requirement. Each such regulation document needs to be reviewed manually by an expert to determine if the document is relevant to the company and, if so, which department it is relevant to. In order to help enterprises improve the efficiency of their operation, we propose an automatic document classification pipeline that determines whether a document is important for the company or not, and if deemed important it forwards those documents to the departments within the company for further review. Binary classification task of determining the importance of a document is done via ensembling the Naive Bayes (NB), support vector machine (SVM), random forest (RF), and artificial neural network (ANN) together for the final prediction, whereas the multi-label classification problem of identifying the relevant departments for a document is executed by the transformer-based DocBERT model. We apply our pipeline to a large corpus of tens of thousands of text data provided by Con Edison and achieve an accuracy score over 80 % . Compared with existing solutions for document classification which rely on a single classifier, our paper i) ensemble multiple classifiers for better accuracy results and escaping from the problem of overfitting, ii) utilize pretrained transformer-based DocBERT model to achieve ideal performance for multi-label classification task and iii) introduce a bi-level structure to improve the performance of the whole pipeline where the binary classification module works as a rough filter before finally distributing the text to corresponding departments through the multi-label classification module. [ABSTRACT FROM AUTHOR]

This paper demonstrates that Q-learning can be used to model Utilitarian decision-making. Accurately modeling ethical theories from the field of moral philosophy is an important step in the development of ethical machine learning. Modeling Utilitarian decision-making with Q-learning is a step toward ethical human–machine teaming; the human and machine contribute according to their strengths to create a more accurate Utilitarian decision than either would make individually. The Utilitarian decision is output by the Q-learning agent, as well as the ranked order of sub-optimal actions to aid in explainability. Additionally, using RL to mathematically represent Utilitarianism solves the classic drawback of Utilitarianism concerning prohibitive computation. This model can also be used for cost-benefit analysis and business decisions. [ABSTRACT FROM AUTHOR]

In this paper we propose a method to augment a Reinforcement Learning agent with legibility. This method is inspired by the literature in Explainable Planning and allows to regularize the agent's policy after training, and without requiring to modify its learning algorithm. This is achieved by evaluating how the agent's optimal policy may produce observations that would make an observer model to infer a wrong policy. In our formulation, the decision boundary introduced by legibility impacts the states in which the agent's policy returns an action that is non-legible because having high likelihood also in other policies. In these cases, a trade-off between such action, and legible/sub-optimal action is made. We tested our method in a grid-world environment highlighting how legibility impacts the agent's optimal policy, and gathered both quantitative and qualitative results. In addition, we discuss how the proposed regularization generalizes over methods functioning with goal-driven policies, because applicable to general policies of which goal-driven policies are a special case. [ABSTRACT FROM AUTHOR]

Under the influence of additive white Gaussian noise, most dictionary learning algorithms whose training data come from the noisy signal cannot effectively remove noise associated with the polynomial phase signal (PPS) via sparse representation, such as K-SVD and RLS-DLA. In this paper, we present a novel dictionary learning algorithm based on neural networks to denoise PPS. In the proposed algorithm, we first use RLS-DLA to train the dictionary. Next, the neural network is used to refine atoms in the learned dictionary. To reduce the computational complexity of iterative calculations in the neural network, direct weight determination of the network is used to denoise atoms. In this way, the signal-to-noise ratio of the reconstructed signal obtained using the proposed algorithm is clearly higher than that of other algorithms, whereas the mean squared error is lower than that of other algorithms. Therefore, the proposed dictionary learning algorithm demonstrates noteworthy denoising performance. [ABSTRACT FROM AUTHOR]

This paper proposes an e-diagnosis system based on machine learning (ML) algorithms to be implemented on the Internet of Medical Things (IoMT) environment, particularly for diagnosing diabetes mellitus (type 2 diabetes). However, the ML applications tend to be mistrusted because of their inability to show the internal decision-making process, resulting in slow uptake by end-users within certain healthcare sectors. This research delineates the use of three interpretable supervised ML models: Naïve Bayes classifier, random forest classifier, and J48 decision tree models to be trained and tested using the Pima Indians diabetes dataset in R programming language. The performance of each algorithm is analyzed to determine the one with the best accuracy, precision, sensitivity, and specificity. An assessment of the decision process is also made to improve the model. It can be concluded that a Naïve Bayes model works well with a more fine-tuned selection of features for binary classification, while random forest works better with more features. [ABSTRACT FROM AUTHOR]

Today, salient object detection has caught the interest of numerous researchers for a variety of applications in computer vision. Most deep learning-based algorithms for SOD tasks produce excellent results but require a lot of data availability and large computational and structural complexities. Also, many methods provide excellent outcomes but are unable to preserve the complete boundaries of the objects in images. The paper discusses an innovative integration method for salient object detection of superpixel segmented images to address these issues. It deals with the integration of saliency maps generated by image decomposition based on non-sub-sampled contourlet transform (NSCT) and by machine learning technique based on the random forest regression using a parameter adaptive pulse coupled neural network (PA-PCNN). The PA-PCNN adaptively estimates each of the free parameters of the pulse-coupled neural network (PCNN). The utilization of PA-PCNN considers neighborhood pixel variances and aids in maintaining object details without fuzziness or distortions. The proposed method restores the edges and boundaries of the objects effectively as PA-PCNN aids in maintaining the perceptually similar attributes of the saliency maps. The results of this study are evaluated using three widely used datasets for detecting salient objects, which show the potential of the proposed system to precisely locate the salient objects in various imaging circumstances like complex background images, images with multiple objects, etc. The quantitative and qualitative experimental results validate a substantial advancement in various evaluation parameters for salient object detection with better boundary preservation of objects. [ABSTRACT FROM AUTHOR]

Phasor Measurement Units (PMU) are capable to generate multi-dimensional time series data, which is one of the most important parts for monitoring power system operation. However, various internal and external factors frequently cause the system to generate anomalous data randomly, so we expect to clean and re-fill the raw PMU data with anomalies to provide support for further advanced applications such as situational awareness, early warning, and dispatch control of the system. Existing methods mostly use classical mathematics and traditional machine learning to analyze PMU data, which makes it difficult to identify the pattern changes of the data under multiple operating conditions in power systems. In this paper, we propose a hybrid model named IRFLMDNN, which consists of an improved CART random forest model and a dynamic neural network optimized by the Levenberg Marquardt algorithm for PMU data anomaly detection and adaptive data re-filling, respectively. Experimental results based on the IEEE 39-node 10-machine New England Power System show that the proposed method has accurate and robust anomaly detection and data refilling performance. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a supervised learning algorithm, which avoids backward recursive gradient computation, for training deep convolutional spiking neural networks (SNNs) with single-spike-based temporal coding. The algorithm employs a linear approximation to compute the derivative of the spike latency with respect to the membrane potential, and it uses spiking neurons with piecewise linear postsynaptic potential to reduce the computational cost and the complexity of neural processing. To evaluate the performance of the proposed algorithm in deep architectures, we employ it in convolutional SNNs for the image classification task. For two popular benchmarks of MNIST and Fashion-MNIST datasets, the network reaches accuracies of, respectively, 99.2 and 92.8 % . The trade-off between memory storage capacity and computational cost with accuracy is analyzed by applying two sets of weights: real-valued weights that are updated in the backward pass and their signs, binary weights, that are employed in the feedforward process. We evaluate the binary CSNN on two datasets of MNIST and Fashion-MNIST and obtain acceptable performance with a negligible accuracy drop with respect to real-valued weights (about 0.6 and 0.8 % drops, respectively). [ABSTRACT FROM AUTHOR]

While many quantum computing techniques for machine learning have been proposed, their performance on real-world datasets remains to be studied. In this paper, we explore how a variational quantum circuit could be integrated into a classical neural network for the problem of detecting pneumonia from chest radiographs. We substitute one layer of a classical convolutional neural network with a variational quantum circuit to create a hybrid neural network. We train both networks on an image dataset containing chest radiographs and benchmark their performance. To mitigate the influence of different sources of randomness in network training, we sample the results over multiple rounds. We show that the hybrid network outperforms the classical network on different performance measures and that these improvements are statistically significant. Our work serves as an experimental demonstration of the potential of quantum computing to significantly improve neural network performance for real-world, non-trivial problems relevant to society and industry. [ABSTRACT FROM AUTHOR]

For semi-supervised learning, a few labeled data and a large number of unlabeled data are used to construct a reasonable classifier. In recent years, many semi-supervised learning methods have been proposed and achieved good performance, especially for the graph-based approaches that can exploit the geometric information embedded in the data. Motivated by the success of generalized elastic net Lp-norm nonparallel support vector machine (GLpNPSVM) and the graph-based regularization term, in this paper, a novel Laplacian generalized elastic net Lp-norm nonparallel support vector machine for semi-supervised learning (Lap-GLpNPSVM) is proposed. A Lp-norm graph regularization term is introduced to improve the performance by the adjustability of the value of p. Experimental results on two synthetic datasets, fifteen UCI datasets, and Handwritten Numeral datasets demonstrate that proposed methods outperform other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

An efficient machine learning-based framework for crop price prediction is proposed in this paper to assist the farmers in estimating their profit-loss beforehand. The proposed work is composed of four functional blocks, such as crop yield prediction, determination of supply, demand prediction and crop price prediction. The input datasets consist of the various field values, such as yield, remaining crop at the end of the year, import, demand and price of a crop. Various time series-based algorithms, such as autoregression, moving average, autoregressive moving average, autoregressive integrated moving average and exponential smoothing, are used to forecast the crop yield. The supply of the crop is determined as a sum of three variables, i.e., the predicted crop yield, residue and import values. The demand for the crop is predicted from a year alone as the demand has more correlation with year over other factors. The crop price from demand, supply and year is predicted using different approaches, which include the time series method, statistical approaches and machine learning techniques. Finally, these three techniques for price prediction are compared to determine the best model having minimum root-mean-square error value. In the proposed work, the decision tree regressor is found to be the best model, for predicting crop price, over others. The superiority of the proposed work over existing approaches, in terms of various aspects, is shown by simulation results. [ABSTRACT FROM AUTHOR]

Automatic facial expression recognition (AFER), sometimes referred to as emotional recognition, is important for socializing. Automatic methods in the past two years faced challenges due to Covid-19 and the vital wearing of a mask. Machine learning techniques tremendously increase the amount of data processed and achieved good results in such AFER to detect emotions; however, those techniques are not designed for masked faces and thus achieved poor recognition. This paper introduces a hybrid convolutional neural network aided by a local binary pattern to extract features in an accurate way, especially for masked faces. The basic seven emotions classified into anger, happiness, sadness, surprise, contempt, disgust, and fear are to be recognized. The proposed method is applied on two datasets: the first represents CK and CK +, while the second represents M-LFW-FER. Obtained results show that emotion recognition with a face mask achieved an accuracy of 70.76% on three emotions. Results are compared to existing techniques and show significant improvement. [ABSTRACT FROM AUTHOR]

The accurate prediction of cardiovascular disease is an essential and challenging task to treat a patient efficiently before occurring a heart attack. In recent times, various intelligent healthcare frameworks have been designed with different machine learning and swarm optimization techniques for cardiovascular disease prediction. However, most of the existing strategies failed to achieve higher accuracy for cardiovascular disease prediction due to the lack of data-recognized techniques and proper prediction methodology. Motivated by the existing challenges, in this paper, we propose an intelligent healthcare framework for predicting cardiovascular heart disease based on Swarm-Artificial Neural Network (Swarm-ANN) strategy. Initially, the proposed Swarm-ANN strategy randomly generates predefined numbers of Neural Networks (NNs) for training and evaluating the framework based on their solution consistency. Additionally, the NN populations are trained by two stages of weight changes and their weight is adjusted by a newly designed heuristic formulation. Finally, the weight of the neurons is modified by sharing the global best weight with other neurons and predicts the accuracy of cardiovascular disease. The proposed Swarm-ANN strategy achieves 95.78% accuracy while predicting the cardiovascular disease of the patients from a benchmark dataset. The simulation results exhibit that the proposed Swarm-ANN strategy outperforms the standard learning techniques in terms of various performance matrices. [ABSTRACT FROM AUTHOR]

Cardiovascular and cerebrovascular diseases are a serious threat to human health and increase the annual death ratio at a considerable pace. This is not uncommon even among teenagers and martial arts athletes. Due to the increased risk associated with strenuous exercise in the context of a quiescent cardiac abnormality, athletes have a higher rate of heart attack and stroke than their nonathletic colleagues. The mortality rate due to this disease is extremely high, which needs to be controlled at the initial stages. At present, the recognition and analysis of ECG signals is still an issue and requires an expert to analyse it and identify its hidden patterns. The most challenging aspect of ECG signal classification is the irregularities in the signals, which are critical for detecting patient status. Each heartbeat is made up of a variety of action impulse waveforms produced by various cardiac tissues. Classification of heartbeats is difficult because waveforms vary from person to person and are identified by certain features. At present, the automatic identification of ECG signals still requires manual design, which has low accuracy and cannot be widely used in clinical practice. This study proposes an intelligent system based on deep learning and machine learning methods to classify and diagnose ECG signals to improve their classification and recognition accuracy. It improves the detection ability of martial arts athletes' arrhythmia disease and obtains accurate arrhythmia diagnosis information. MIT-BIH arrhythmia dataset has been used for the experimental analysis. The performance of the proposed scheme is evaluated with the help of various performance measures. We conduct comprehensive experiments, and the results show that the algorithms used in this paper are robust. [ABSTRACT FROM AUTHOR]

Hypertension is a primary or contributing cause for premature death in the entire world. As a matter of fact, there is a high prevalence and low control rates in low- and middle-income countries, such that the prevention and treatment of hypertension should remain a top priority in global health. In the recent years, the awareness, treatment, and control rates of hypertension patients in China have been significantly improved to 51.6%, 45.8%, and 16.8%, respectively. However, those rates are still far from a satisfactory level. Clinical studies suggest that for people in the pre-clinical stage of hypertension or having the risk of hypertension, the progression of the disease may be significanly reduced through a change in lifestyle, or by an effective drug therapy. In this paper, we address risk prediction for hypertension in the next five years, and put forward a model merging KNN and LightGBM. Our approach allows us to predict the hypertension risk for a specific individual using features such as the age of the subject and blood indicators. Results shows that our model is reliable and achieves accuracy and recall rate over 86% and 92%, respectively. [ABSTRACT FROM AUTHOR]

In sports, advance sensing technologies generate massive amount of unstructured telemedicine data that need to be refined for accurate diagnosis of underlying diseases. For accurate prediction of diseases and classification of athletes' data, deep learning algorithms are frequently used at the cloud. However, the transmission of raw data of athletes to the cloud faces numerous challenges. Among them, security and privacy are a major challenge in view of the sensitive and personal information present within the unstructured data. In this paper, first we present a data block scrambling algorithm (without key management) for secured transmission and storage of ECG (electrocardiogram) data of table tennis players at the cloud. A small piece of original data stored at the cloud is used for scrambling the massive amount of remaining ECG data. The secured telemedicine data is then imported into Hadoop Distributed File System for data management, which is read by Spark framework to form Resilient Distributed Datasets. Finally, a deep learning approach is used that extracts useful features, learns the related information, and weights and sums the feature vectors at different layers for classification. Theoretical analysis proves that our proposed approach is highly robust and resilient to brute force attacks and at the same time has a much better accuracy, sensitivity, and specificity as compared to the existing approaches. [ABSTRACT FROM AUTHOR]

Medical data processing and analysis using machine learning algorithms are a prominent research topic these days. To obtain high performances, most state-of-the-art models must be trained on a large number of labeled datasets. However, manually collecting a large-scale real-world medical data is expensive due to the issues like privacy, security, and reliability. Moreover, the collected samples may be incomplete, i.e., some important data items are missing, which has a significant impact on the performance of the machine learning algorithms, especially deep learning models. In this paper, we present a strategy that uses the idea of adversarial learning to augment the real-world medical samples with the incomplete issue to obtain better performance in predicting patient states. Rather than supplying the incomplete data sample with extra instance-level information as in existing data-complementation methods, our method aims to achieve the complementary effect in the feature level without consuming human effort. The method receives both high-quality data and low-quality data (with serious incomplete issue) to learn a comprehensive feature space where the incomplete samples can be complemented with the knowledge transferred from the high-quality samples by the adversarial learning scheme. From the perspective of feature representations, our method can alleviate the incomplete issue of the low-quality data, which enhances the model performance in the end task. Experiments show that our method works well on a real-world medical dataset, which collected for assessment of pelvic organ prolapse. When compared to machine learning approaches frequently employed in medical data analysis, our approach shows a significant improvement of up to 10%. [ABSTRACT FROM AUTHOR]

The coronavirus pandemic has been globally impacting the health and prosperity of people. A persistent increase in the number of positive cases has boost the stress among governments across the globe. There is a need of approach which gives more accurate predictions of outbreak. This paper presents a novel approach called diffusion prediction model for prediction of number of coronavirus cases in four countries: India, France, China and Nepal. Diffusion prediction model works on the diffusion process of the human contact. Model considers two forms of spread: when the spread takes time after infecting one person and when the spread is immediate after infecting one person. It makes the proposed model different over other state-of-the art models. It is giving more accurate results than other state-of-the art models. The proposed diffusion prediction model forecasts the number of new cases expected to occur in next 4 weeks. The model has predicted the number of confirmed cases, recovered cases, deaths and active cases. The model can facilitate government to be well prepared for any abrupt rise in this pandemic. The performance is evaluated in terms of accuracy and error rate and compared with the prediction results of support vector machine, logistic regression model and convolution neural network. The results prove the efficiency of the proposed model. [ABSTRACT FROM AUTHOR]

In an iron ore pelletization plant, pellets are produced inside a rotating disc pelletizer. Online pellet size distribution is an important performance indicator of the pelletization process. Image processing-based system is an effective solution for online size analysis of iron ore pellets. This paper proposes a machine learning algorithm for estimating the size class of the pellets during their production by imaging from an area inside the disc pelletizer. Instead of computing the size of each individual pellets in the acquired image, this method proposes a qualitative approach to get the overall size estimate of the pellets in production. The key idea of this paper is to find out whether the disc is producing VERY SMALL, SMALL, MEDIUM, or BIG-sized pellets. A weighted average ensemble of different convolutional neural networks such as VGG16, Mobilenet, and Resnet50 is used to achieve this objective. Furthermore, batch normalization is applied to improve the estimation performance of the proposed model. A novel data augmentation method is applied to the in situ captured images to create the data set used to train and evaluate the proposed ensemble of CNN models. Results of experiments indicate that it is possible to detect the operating state of the pelletization disc by acquiring images from the inside area of the disc with sufficient accuracy. [ABSTRACT FROM AUTHOR]

This document is a special issue of the journal "Neural Computing & Applications" focused on machine learning and big data analytics for IoT security and privacy. The Internet of Things (IoT) is a network of connected devices that presents challenges for security and analysis due to the large amounts of data involved. The issue includes selected papers from the 3rd International Conference on Machine Learning and Big Data Analytics for IoT Security and Privacy (SPIoT2022), as well as additional submissions. The papers cover a range of topics including prototype selection, scheduling optimization, deep learning, text reading, task-technology matching, financial incentives, state estimation for robots, image recognition, speech recognition, and term extraction algorithms. The authors express gratitude to the editor, staff, authors, and reviewers involved in the creation of this issue. [Extracted from the article]

Recently, the COVID-19 epidemic has resulted in millions of deaths and has impacted practically every area of human life. Several machine learning (ML) approaches are employed in the medical field in many applications, including detecting and monitoring patients, notably in COVID-19 management. Different medical imaging systems, such as computed tomography (CT) and X-ray, offer ML an excellent platform for combating the pandemic. Because of this need, a significant quantity of study has been carried out; thus, in this work, we employed a systematic literature review (SLR) to cover all aspects of outcomes from related papers. Imaging methods, survival analysis, forecasting, economic and geographical issues, monitoring methods, medication development, and hybrid apps are the seven key uses of applications employed in the COVID-19 pandemic. Conventional neural networks (CNNs), long short-term memory networks (LSTM), recurrent neural networks (RNNs), generative adversarial networks (GANs), autoencoders, random forest, and other ML techniques are frequently used in such scenarios. Next, cutting-edge applications related to ML techniques for pandemic medical issues are discussed. Various problems and challenges linked with ML applications for this pandemic were reviewed. It is expected that additional research will be conducted in the upcoming to limit the spread and catastrophe management. According to the data, most papers are evaluated mainly on characteristics such as flexibility and accuracy, while other factors such as safety are overlooked. Also, Keras was the most often used library in the research studied, accounting for 24.4 percent of the time. Furthermore, medical imaging systems are employed for diagnostic reasons in 20.4 percent of applications. [ABSTRACT FROM AUTHOR]

Credit risk assessment is at the core of modern economies. Traditionally, it is measured by statistical methods and manual auditing. Recent advances in financial artificial intelligence stemmed from a new wave of machine learning (ML)-driven credit risk models that gained tremendous attention from both industry and academia. In this paper, we systematically review a series of major research contributions (76 papers) over the past eight years using statistical, machine learning and deep learning techniques to address the problems of credit risk. Specifically, we propose a novel classification methodology for ML-driven credit risk algorithms and their performance ranking using public datasets. We further discuss the challenges including data imbalance, dataset inconsistency, model transparency, and inadequate utilization of deep learning models. The results of our review show that: 1) most deep learning models outperform classic machine learning and statistical algorithms in credit risk estimation, and 2) ensemble methods provide higher accuracy compared with single models. Finally, we present summary tables in terms of datasets and proposed models. [ABSTRACT FROM AUTHOR]

In order to improve the effect of e-government work, it is necessary to analyze its influencing factors. E-government is affected by many social factors, so it is necessary to combine intelligent models to improve the effect of factor analysis. This paper combines the essence of e-government influencing factor data to improve the machine learning algorithm and uses the EM algorithm to derive the parameter estimation formula of the data in the case of missing data to improve the accuracy of data analysis. Moreover, this article combines the structure of the e-government system to build the main structure of the intelligent analysis model of the influence factors of e-government. According to the key influencing factor model of e-government adoption and multi-dimensional research on technology, organization and implementation, this paper puts forward the model and promotion mode of the e-government system based on cloud computing and conducts a simulation. From the simulation results, the effect of the model proposed in this paper is more significant. [ABSTRACT FROM AUTHOR]

To improve the effect of landslide feature extraction, this paper improves the remote sensing image recognition algorithm with the support of a machine learning algorithm. Moreover, this paper combines UAV remote sensing images to extract landslide features, classifies and introduces the evaluation criteria for target detection and several representative target detectors. This paper also constructs the functional structure of the system according to the landslide feature extraction requirements and designs a set of optimization schemes for landslide feature data collection and control measurement suitable for field operations. In addition, this paper analyses the system kernel algorithm process and analyses the system function realization through simulation research. Finally, this paper designs an experiment to evaluate the practicability of the system constructed in this paper. From the results of experimental statistics, we can see that the system constructed in this paper has good practicability. [ABSTRACT FROM AUTHOR]

The development of green finance and the quantitative evaluation of its impact on the ecological environment provide empirical evidence for the construction of the carbon trading accounting system. Among them, carbon trading is an important part of green finance, and the accounting of businesses related to carbon emission rights has promoted the development of regional green finance. In order to explore the relationship between green finance and carbon emissions, this paper builds an analysis model of the relationship between green finance and carbon emissions based on big data and machine learning based on big data technology and machine learning technology. Moreover, this paper conducts simulation tests through the system and compares the output results with the actual situation after system simulation to verify the effectiveness of the model in this paper. From the experimental research results, it can be seen that the correlation analysis model of green finance and carbon emissions based on big data and machine learning constructed in this paper has a good performance in the correlation analysis of green finance and carbon emissions. Moreover, it is not difficult to see through the model of this paper that there is a clear correlation between green finance and carbon emissions. [ABSTRACT FROM AUTHOR]

The heterogeneity of enterprise management will have a certain impact on enterprise investment. In order to explore the strength of this correlation, this paper builds an intelligent system that can be used for enterprise analysis with the support of machine learning technology, uses embedded algorithms to improve the traditional algorithm structure, and combines machine learning to optimize and analyze data processing. Moreover, this paper combines the heterogeneity of enterprise management and the actual needs of enterprise management, the current financial market investment status to analyze the system's functional structure, and combine the actual needs to construct the overall system architecture. In addition, this paper adopts algorithm analysis to the system process, analyzes the system logic layer structure, and builds the overall system structure framework. Finally, this paper designs an experiment to analyze the performance of the system constructed in this paper. From the experimental research, it can be seen that the system constructed in this paper basically meets the expected requirements. [ABSTRACT FROM AUTHOR]

People's cognition of objectively existing things is developing in the direction of digitization, and more and more objectively existing things are constantly being presented in the form of data. In order to realize the analysis of constantly changing large-scale data and the automatic extraction of valuable information, this paper analyzes the application of artificial intelligence auxiliary equipment based on machine learning in the cognitive learning process, and proposes a collaborative filtering method based on fusion of global and local parameters. Moreover, this paper studies a cognitive computing model based on context-aware data stream to realize effective analysis of context-aware data and obtain effective cognitive results. In addition, this paper proposes a task scheduling algorithm based on improved queue matching to solve the task scheduling problem in a distributed computing environment. Finally, this paper constructs an improved algorithmic cognitive system architecture, and verifies the performance of the system through experimental research. The research results show that the system constructed in this paper is highly reliable. [ABSTRACT FROM AUTHOR]

Due to the uncertainty of geological conditions during the tunneling process, advanced prediction of TBM tunneling parameters is significant for evaluating operational safety and efficiency, especially for real-time prediction of key tunneling parameters during the steady phase of TBM operation. At present, although there are studies on constructing predictive models based on machine learning algorithms, multiparameter prediction consistent with the actual tunneling process remains challenging due to the complexity of the TBM tunneling process and the numerous tunneling parameters. Therefore, this paper proposes a real-time multiple tunneling parameters prediction method of TBM steady phase based on dual recurrent neural networks. Firstly, the irregular multidimensional time series of tunneling parameters are analyzed and processed, which are divided into an idle-push phase, a rising phase, and a steady phase; secondly, the parameters of rising phase are analyzed using a recurrent neural network, and the parameters relevant for constructing a real-time prediction model are screened; then, based on the screened parameters, the Bayesian-optimized gated recurrent unit (GRU, a kind of recursive neural network) is proposed to construct a real-time prediction model for the four key tunneling parameters during the steady phase. Finally, the effectiveness and practicality of the proposed method are demonstrated by verification on real TBM tunnel datasets and comparing it with the models constructed by six commonly used machine learning algorithms. The results of this paper show that the designed prediction method is able to achieve a good combination of performance in terms of accuracy and computational time-consumption, with an average prediction accuracy of 91.1% for the four parameters for different rock grades of geology, the multiparameter prediction time for 100 samples is only 11 ms. In addition, three current similar studies using deep learning methods were compared to demonstrate the superiority of this proposal. As a method more closer to practical application, this work provides guidance for the forward-looking prediction of TBM tunneling parameters. [ABSTRACT FROM AUTHOR]

POI group recommendation is one of the hottest research topics in location-based social networks, which recommends the most agreeable places for a group of users. However, traditional POI group recommendation methods only generate a consensus function to aggregate individual preference into group preference and they do not consider all the factors that can determine the results of POI group recommendation, which leads to a low recommendation accuracy. What's more, these methods have a long running time. Therefore, in this paper, we propose a new POI group recommendation method with an extreme learning machine (ELM) called PGR-ELM. The PGR-ELM method regards POI group recommendation as a binary classification problem. First, three features are extracted from three factors: POI popularity, group members' distance to POI, members' interest preferences combined affinity between group members. These features simultaneously consider all the factors that can determine the results of recommendation and guarantee the effectiveness of POI group recommendation. Then, the extracted features are input to train an ELM classifier because of its fast learning speed, which guarantees the efficiency of POI group recommendation. Finally, extensive experiments verify the accuracy and efficiency of PGR-ELM method. [ABSTRACT FROM AUTHOR]

Stress is now thought to be a major cause to a wide range of human health issues. However, many people may ignore their stress feelings and disregard to take action before serious physiological and mental disorders take place. The heart rate (HR) and blood pressure (BP) are the most physiological markers used in various studies to detect mental stress for a human, and because they are captured non-invasively using wearable sensors, these markers are recommended to provide information on a person's mental state. Most stress assessment studies have been undertaken in a laboratory-based controlled environment. This paper proposes an approach to identify the mental stress of automotive drivers based on selected biosignals, namely, ECG, EMG, GSR, and respiration rate. In this study, six different machine learning models (KNN, SVM, DT, LR, RF, and MLP) have been used to classify between the stressed and relaxation states. Such system can be integrated with a Driver Assistance System (DAS). The proposed stress detection technique (SDT) consists of three main phases: (1) Biosignal Pre-processing, in which the signal is segmented and filtered. (2) Feature Extraction, in which some discriminate features are extracted from each biosignal to describe the mental state of the driver. (3) Classification. The results show that the RF classifier outperforms other techniques with a classification accuracy of 98.2%, sensitivity 97%, and specificity 100% using the drivedb dataset. [ABSTRACT FROM AUTHOR]

Diabetes Mellitus (DM) is a widespread condition that is one of the main causes of health disasters around the world, and health monitoring is one of the sustainable development topics. Currently, the Internet of Things (IoT) and Machine Learning (ML) technologies work together to provide a reliable method of monitoring and predicting Diabetes Mellitus. In this paper, we present the performance of a model for patient real-time data collection that employs the Hybrid Enhanced Adaptive Data Rate (HEADR) algorithm for the Long-Range (LoRa) protocol of the IoT. On the Contiki Cooja simulator, the LoRa protocol's performance is measured in terms of high dissemination and dynamic data transmission range allocation. Furthermore, by employing classification methods for the detection of diabetes severity levels on acquired data via the LoRa (HEADR) protocol, Machine Learning prediction takes place. For prediction, a variety of Machine Learning classifiers are employed, and the final results are compared with the already existing models where the Random Forest and Decision Tree classifiers outperform the others in terms of precision, recall, F-measure, and receiver operating curve (ROC) in the Python programming language. We also discovered that using k-fold cross-validation on k-neighbors, Logistic regression (LR), and Gaussian Nave Bayes (GNB) classifiers boosted the accuracy. [ABSTRACT FROM AUTHOR]

Computations that mimic nature are known as nature-inspired computing. Nature presents a wealthy source of thoughts and ideas for computing. The use of natural galvanized techniques has been found to provide machine solutions to complex problems. One of the challenging issues among researchers is high-dimensional data which contains a large number of unwanted, redundant, and irrelevant features. These redundant or unwanted features reduce the accuracy of machine learning models. Therefore, to solve this problem nowadays metaheuristic techniques are being used. The paper presents both surveys as well as comparison of five metaheuristic algorithms for feature selection. A wrapper-based feature selection approach using five nature-inspired techniques for feature selection has been applied. The binary version of the five swarm-based nature-inspired algorithms (NIAs), namely particle swarm optimization, whale optimization algorithm (WOA), grey wolf optimization (GWO), firefly algorithm, and bat algorithm. WOA and GWO are recent algorithms used for finding optimal feature subsets when there is no empirical information. The S-shape transfer function has been used to convert the continuous value to binary form and K-nearest neighbor is used to calculate the classification accuracy of selected feature subsets. To validate the results of the selected NIAs eleven benchmark datasets from the UCI repository are used. The strength of each NIA has been verified using a nonparametric test called the Friedman rank and Holm test. p value obtained shows that WOA is statistically significant and performs better than other models. [ABSTRACT FROM AUTHOR]

Human Activity Recognition (HAR) based on Channel State Information (CSI) has important application prospects in various fields such as human–computer interactivity, medical health et al. Although the current CSI-based HAR researches have made great progress in the number of activity categories and recognition accuracy, they encounter two challenges. When recognizing new activities, a large number of new activity samples are required to retrain the original model or adjust the model parameters. Besides, the recognition accuracy of the model retrained in the new scene for the already recognized activities drops obviously. To address the two challenges, we propose a human activity recognition system CSI-ARIL based on incremental learning in this paper. CSI-ARIL saves partial representative samples for each learned activity, and adds distillation loss to the loss function as the regularization term to further strengthen the model's memory for the learned activities. In order to enhance the feature difference among different activities, CSI-ARIL explicitly adds time information for CSI activity samples. In addition, pseudo-samples are generated through data enhancement to expand the new training set, which reduces the cost of obtaining labeled real samples. CSI-ARIL also adds the Convolutional Block Attention Module (CBAM) to the network, which helps the network extract the effective features of CSI activity samples. The experimental results show that, after continuously learning eight categories of activities in four scenes, the recognition accuracy of CSI-ARIL for new and already recognized activities reaches 92.8% and 89.4%, respectively. [ABSTRACT FROM AUTHOR]

The heat exchanger has been widely used in the energy and chemical industry and plays an irreplaceable role in the featured applications. The design of heat exchanger is a mixed integer complex optimization problem, where the efficient design significantly improves the efficiency and reduces the cost. Many intelligent methods have been developed for heat exchanger optimal design. In this paper, a novel variant of sine and cosine algorithm named EDOLSCA is proposed, enhanced by dynamic opposite learning algorithm and the elite strategy. The proposed method is tested in CEC2014 benchmark and proved to be of significant advantages over the original algorithm. The new algorithm is then validated in the plate-fin heat exchanger (PFHE) optimal design problem. The comparison results of the proposed algorithm and other algorithms prove that EDOLSCA also has demonstrated superiority in heat exchanger optimal design. [ABSTRACT FROM AUTHOR]

Mobile and edge computing devices for always-on classification tasks require energy-efficient neural network architectures. In this paper we present several changes to neural architecture searches that improve the chance of success in practical situations. Our search simultaneously optimizes for network accuracy, energy efficiency and memory usage. We benchmark the performance of our search on real hardware, but since running thousands of tests with real hardware is difficult, we use a random forest model to roughly predict the energy usage of a candidate network. We present a search strategy that uses both Bayesian and regularized evolutionary search with particle swarms, and employs early stopping to reduce the computational burden. Our search, evaluated on a sound event classification dataset based upon AudioSet, results in an order of magnitude less energy per inference and a much smaller memory footprint than our baseline MobileNetV1/V2 implementations while slightly improving task accuracy. We also demonstrate how combining a 2D spectrogram with a convolution with many filters causes a computational bottleneck for audio classification and that alternative approaches reduce the computational burden but sacrifice task accuracy. [ABSTRACT FROM AUTHOR]

One of the main indicators for detecting changes in climate and marine ecosystems around the world is the sea surface temperature (SST). Even with several models presented in the literature, it is still a challenging task when only a single model is considered for SST forecasting. In this context, hybrid approaches that combine statistical models (to estimate linear dependencies) and machine learning models (to estimate nonlinear dependencies from residuals) have attained highlighted accuracy in several time series forecasting problems. In this way, this paper proposes a hybrid system that combines the autoregressive integrated moving average (ARIMA) model with a deep morphological neural network model, which employs dilation-erosion operators to estimate ARIMA's model residuals. In this context, the proposed hybrid system is composed of three stages: (1) time series forecast using the ARIMA model, (2) residuals forecast using the deep morphological neural network and (3) a linear combination of the stages (1) and (2). Three SST time series are used in the experimental analysis. Moreover, the achieved results are evaluated using three relevant statistical measures showing that the proposed hybrid system attains higher accuracy when compared to other hybrid models in the literature. [ABSTRACT FROM AUTHOR]

Around 3 million people worldwide have an arm amputation. These people face a lot of trouble in their everyday life whilst performing basic tasks. This paper proposes a novel deep learning-based approach for predicting arm movements using EEG-based signals. We plan to design and develop an active exoskeleton controlled by the same EEG-based signals to rehabilitate the amputees. The architecture design is intended to build an exoskeleton arm with at least 3 degrees of freedom that can perform complex movements and is sophisticated enough to substitute for a real arm. This prosthetic arm will be controlled using electroencephalogram (EEG) signals gathered by different devices/headsets and processed using deep learning models. The results show that our proposed approach gives excellent results. [ABSTRACT FROM AUTHOR]

Rib segmentation based on chest X-ray images is essential in the computer-aided diagnosis systems of lung cancer, which serves as an important step in the quantitative analysis of various types of lung diseases. However, the traditional methods are unable to segment ribs effectively due to the unclear edges and overlapping regions in X-ray images. A novel rib segmentation framework based on Unpaired Sample Augmentation and Multi-Scale Network is presented in this paper, aiming to improve the accuracy of ribs segmentation with limited labeled samples. First, the algorithm learns pneumonia-related texture changes via unpaired chest x-ray images and generates various augmented samples. Then, a multi-scale network attempts to learn hierarchical features using global supervision. Finally, the refined segmentation result of each organ is achieved by using a deep separation module and a comprehensive loss function. Specifically, the hierarchical features can greatly improve the robustness of multi-organ segmentation networks. The complex multi-organ segmentation task with limited labeled data is simplified with the designed deep separation module. We justify the proposed framework through extensive experiments. It achieves good performance with DSC, Precision, Recall, and Jaccard of 88.03, 88.25, 88.36, and 79.02%, respectively. The DSC value increases nearly by 3% compared to other popular methods. The experimental results show that our algorithm presents better segmentation performance for the overlapping region and fuzzy region of multiple organs, which holds research value and prospects for application. [ABSTRACT FROM AUTHOR]

Recently, the COVID-19 pandemic has triggered different behaviors in education, especially during the lockdown, to contain the virus outbreak in the world. As a result, educational institutions worldwide are currently using online learning platforms to maintain their education presence. This research paper introduces and examines a dataset, E-LearningDJUST, that represents a sample of the student's study progress during the pandemic at Jordan University of Science and Technology (JUST). The dataset depicts a sample of the university's students as it includes 9,246 students from 11 faculties taking four courses in spring 2020, summer 2020, and fall 2021 semesters. To the best of our knowledge, it is the first collected dataset that reflects the students' study progress within a Jordanian institute using e-learning system records. One of this work's key findings is observing a high correlation between e-learning events and the final grades out of 100. Thus, the E-LearningDJUST dataset has been experimented with two robust machine learning models (Random Forest and XGBoost) and one simple deep learning model (Feed Forward Neural Network) to predict students' performances. Using RMSE as the primary evaluation criteria, the RMSE values range between 7 and 17. Among the other main findings, the application of feature selection with the random forest leads to better prediction results for all courses as the RMSE difference ranges between (0–0.20). Finally, a comparison study examined students' grades before and after the Coronavirus pandemic to understand how it impacted their grades. A high success rate has been observed during the pandemic compared to what it was before, and this is expected because the exams were online. However, the proportion of students with high marks remained similar to that of pre-pandemic courses. [ABSTRACT FROM AUTHOR]

Data imbalance is one of the most difficult problems in machine learning. The improved ensemble learning model is a promising solution to mitigate this challenge. In this paper, an improved multi-class imbalanced data classification framework is proposed by combining the Focal Loss with Boosting model (FL-Boosting). By addressing the confusion of the second-order derivation of Focal Loss in the traditional ensemble learning model, the proposed model achieves a more efficient and accurate classification of the imbalanced data. More specifically, a Highly Adaptive Focal Loss (HAFL) is proposed to ensure that the model maintains lasting attention to the minority samples, which could be combined with boosting model to build HAFL-Boosting to achieve better performance. The framework has the scalability to adapt to different situations according to typical ensemble learning algorithms such as LightGBM, XGBoost and CatBoost. In addition, to implement the application of the proposed framework on deep models, a two-stage classification method combining ConvNeXt with the improved boosting model is proposed, which could improve the recognition ability to high-dimensional imbalanced data. We evaluate the HAFL-Boosting and the two-stage class imbalance classification method by ablation experiments and benchmark experiments, which demonstrated that the proposed methods obviously improved the scores on several evaluation indexes. The comparative experiments with the latest classification models show that the proposed methods could achieve leading performance from multiple perspectives. [ABSTRACT FROM AUTHOR]

Manual classification of millions of songs of the same or different genres is a challenging task for human beings. Therefore, there should be a machine intelligent model that can classify the genres of the songs very accurately. In this paper, a deep learning-based hybrid model is proposed for the analysis and classification of different music genre files. The proposed hybrid model mainly uses a combination of multimodal and transfer learning-based models for classification. This model is analyzed using GTZAN and Ballroom datasets. The GTZAN dataset contains 1000 music files classified with 10 different kinds of music genres such as Metal, Classical, Rock, Reggae, Pop, Disco, Blues, Country, Hip-Hop and Jazz, and the duration of each music file is 30 s. The Ballroom dataset contains 698 music files classified into 8 different kinds of music genres such as Tango, ChaChaCha, Rumba, Viennese waltz, Jlive, Waltz, Quickstep and Samba, and the duration of each music file is 30 s. The performance of the model is evaluated using the Python tool. The macro-average and weighted average are taken for computing the percentage of accuracy of each model. From the results, it is found that the proposed hybrid model is able to perform better as compared to other deep learning models such as the convolution neural network model, transfer learning-based model, multimodal model, machine learning models and other existing models in terms of training accuracy, validation accuracy, training loss, validation loss, precision, recall, F1-score and support. [ABSTRACT FROM AUTHOR]

In the previous image segmentation methods based on affinity graph learning, it is difficult to obtain clear contour boundaries in the process of image preprocessing, over-segmentation leads to the separation of local regions and most traditional affinity graph learning methods cluster and segment with a single feature of natural images, ignoring the associative characteristics of multiple types of features and cannot effectively utilize the useful information of multiple features of natural images. Aiming at such problems, this paper proposes a tensor-based multi-feature affinity graph learning method for natural image segmentation (TRMFAL). First, the adaptive morphological rebuilding watershed transform is applied to original natural image, the obtained superpixel image contains a lot of boundary contour information, and the fusion of local regions is relatively good; secondly, extract multi-class features from the superpixel blocks in the superpixel image, effectively utilize the features of different characteristics, and integrate them into a multi-feature data matrix according to the corresponding rules; Then, a tensor-based multi-feature affinity graph learning algorithm is proposed, in which tensor is introduced in the algorithm to effectively obtain the higher-order information of image data, and use the projection matrix to embed the original data in the low-dimensional space to decrease the dimension, while minimizing the residual error of each view feature and assign appropriate weights according to the importance of the feature information; finally, use spectral clustering performs clustering and segmentation on affinity graphs to obtain final clustering results and segmented images. In addition, an optimization iterative algorithm based on the alternating multiplier direction method is designed, which effectively solves the problem of solving the TRMFAL model. Sufficient experimental comparisons are carried out on multiple public datasets, and the results prove that the proposed method achieves the optimal clustering performance and segmentation effect. [ABSTRACT FROM AUTHOR]

In the few-shot class-incremental learning, new class samples are utilized to learn the characteristics of new classes, while old class exemplars are used to avoid old knowledge forgetting. The limited number of new class samples is more likely to cause overfitting during incremental training. Moreover, mass stored old exemplars mean large storage space consumption. To solve the above difficulties, in this paper we propose a novel flexible few-shot class-incremental framework to make the incremental process efficient and convenient. We enhance the expression ability of extracted features through multistage pre-training. Then, we set up a prototype container to store each class prototype to retain old knowledge. When new classes flow in, we calculate the new class prototypes and update the prototype container. Finally, we get the prediction result through similarity weighting. The entire framework only need to train the base class classifier and does not require further training during the incremental process. It avoids the overfitting of novel classes and saves time for further training. Besides, storing prototypes can save more storage space than original image data. Overall, the entire framework has the advantage of flexibility. We conduct extensive experiments on three standard few-shot class-incremental datasets and achieve state-of-the-art results. Especially, to verify the flexibility of the framework, we discuss the special federated few-shot class-incremental scenarios in addition. No further training and less storage consumption provide the possibility for applications in more complex scenarios. [ABSTRACT FROM AUTHOR]

In multimodal meta-learning, previous works modulate the meta-learned network to adapt to tasks, based on task embeddings extracted by an task encoder. However, it is ignored that similarity exists in the tasks from the same mode, and the similarity cannot be utilized when generating the initialization and loss for the task-learner in previous methods. In this paper, we propose a new method to leverage the task similarity in multimodal meta-learning, which provides a better-suited initialization and loss for the task-learner with consideration of the task characteristics. In our proposed Task Embedding Adaptation (TEA), a Transformer is introduced to refine the task embeddings by encouraging information aggregation among similar tasks. The enhanced task embeddings can be utilized to infer the task mode more accurately and modulate the meta-learner to generate a better task-specific initialization. Furthermore, Modulated Adaptive Loss module is proposed to generate task-specific loss adaptively based on the loss network and enhanced task embeddings obtained by TEA. In addition, we present an efficient Mixed-Loop Learning strategy to ensure the training efficiency of the task encoder and loss network instead of the traditional two-loop learning strategy. Extensive experiments on multimodal few-shot classification demonstrate that our method achieves state-of-the-art performance, and our method also performs well in regression and reinforcement learning. [ABSTRACT FROM AUTHOR]

Emergency medicine (EM) is one of the attractive research fields in which researchers investigate their efforts to diagnose and treat unforeseen illnesses or injuries. There are many tests and observations are involved in EM. Detection of the level of consciousness is one of these observations, which can be detected using several methods. Among these methods, the automatic estimation of the Glasgow coma scale (GCS) is studied in this paper. The GCS is a medical score used to describe a patient's level of consciousness. This type of scoring system requires medical examination that may not be available with the shortage of the medical expert. Therefore, the automatic medical calculation for a patient's level of consciousness is highly needed. Artificial intelligence has been deployed in several applications and appears to have a high performance regarding providing automatic solutions. The main objective of this work is to introduce the edge/cloud system to improve the efficiency of the consciousness measurement through efficient local data processing. Moreover, an efficient machine learning (ML) model to predict the level of consciousness of a certain patient based on the patient's demographic, vital signs, and laboratory tests is proposed, as well as maintaining the explainability issue using Shapley additive explanations (SHAP) that provides natural language explanation in a form that helps the medical expert to understand the final prediction. The developed ML model is validated using vital signs and laboratory tests extracted from the MIMIC III dataset, and it achieves superior performance (mean absolute error (MAE) = 0.269, mean square error (MSE) = 0.625, R2 score = 0.964). The resulting model is accurate, medically intuitive, and trustworthy. [ABSTRACT FROM AUTHOR]

In the last decades, there has been increasing awareness of the different types of inequalities that women experience. A very important inequality is the wage gap. Understanding the elements that affect this gap is crucial in order for governments to take the right actions to diminish the gap. It is also important to understand the broader context in which this inequality has evolved over time. In this paper, we develop a causal inference model based on the ideas of Potential Outcome (PO) and Metalearners (ML) to address this important issue. We include a time variable in the causal analysis which helps to determine how the effects have evolved over the last decades. We apply data from 1990 to 2017 from the official government social survey of Chile to fit the models. We then make a deep analysis of each variable using the SHAP framework to see the impact of each variable on the gender wage gap. Sadly, our results indicate that there has been a gap between the earnings of men and women over the last three decades, and the gap actually widened over time. We also find that variable decomposition helps to clarify the different effects as some variables clearly help to diminish this gap. Our results may assist the government of Chile and other organizations to endorse policies that may reduce the gap. [ABSTRACT FROM AUTHOR]

Class imbalance learning deals with data that have very skewed class distributions, and commonly exists in real-world applications. Incremental learning has the ability to train a model continually using new data, which requires the model to learn the new information without forgetting the old one. While these two issues have been independently discussed, their joint treatment has not been studied thoroughly. This paper studies the combined challenges and proposes a balanced self-organizing incremental neural network (Balanced SOINN). First, we introduce Balanced SOINN, which can be trained incrementally with the resampling method. Then, we compare Balanced SOINN with other methods with artificial and real-world data. Our proposed method is competitive in artificial datasets in non-incremental scenarios and achieves the best performance with real-world datasets in incremental scenarios. [ABSTRACT FROM AUTHOR]

Container-based cloud technology has changed the delivery mode of traditional applications and brought a breakthrough development to the field of cloud computing. However, the uncertainty of cloud environment and variability of application requirements increase the scheduling cost of tasks in container cloud. In particular, how to balance the business performance and utilization efficiency of cloud resources in the peak stage of application access is the focus of future for container cluster technology. In this paper, we propose a heuristics multi-objective task scheduling framework based on reinforcement learning (AC-CCTS). The proposed framework not only solves the problems of single objective and local convergence in traditional task scheduling methods, but also reduces the cost of experiential learning with reinforcement learning methods. Firstly, we define container cloud environment, scheduling agent, scheduling actions and scheduling evaluation methods to establish a deep reinforcement learning-based dynamic scheduling model. Then, based on Actor-Critic algorithm, we design heuristic rules and prioritized experience replay method to speed up convergence of task scheduling and decrease learning costs. At the same time, we provide compensation mechanism for dynamic task scheduling to improve the robustness of the approach. Finally, we implement comparative experiments to simulate various scheduling scenarios and verify the effectiveness of AC-CCTS from different perspectives such as resource balance, resource utilization and QoS. Compared with traditional meta-heuristic scheduling methods such as FIMPSO, HWOA-MBA and other reinforcement learning algorithms such as DeepRM-Plus and RLSched, AC-CCTS shows better resource utilization efficiency and convergence stability in container-based cloud task scheduling. [ABSTRACT FROM AUTHOR]