Search

The spread of Covid-19 misinformation on social media had significant real-world consequences, and it raised fears among internet users since the pandemic has begun. Researchers from all over the world have shown an interest in developing deception classification methods to reduce the issue. Despite numerous obstacles that can thwart the efforts, the researchers aim to create an automated, stable, accurate, and effective mechanism for misinformation classification. In this paper, a systematic literature review is conducted to analyse the state-of-the-art related to the classification of misinformation on social media. IEEE Xplore, SpringerLink, ScienceDirect, Scopus, Taylor & Francis, Wiley, Google Scholar are used as databases to find relevant papers since 2018–2021. Firstly, the study begins by reviewing the history of the issues surrounding Covid-19 misinformation and its effects on social media users. Secondly, various neuro-fuzzy and neural network classification methods are identified. Thirdly, the strength, limitations, and challenges of neuro-fuzzy and neural network approaches are verified for the classification misinformation specially in case of Covid-19. Finally, the most efficient hybrid method of neuro-fuzzy and neural networks in terms of performance accuracy is discovered. This study is wrapped up by suggesting a hybrid ANFIS-DNN model for improving Covid-19 misinformation classification. The results of this study can be served as a roadmap for future research on misinformation classification. [ABSTRACT FROM AUTHOR]

The popularity of digital cameras and social networks has greatly enriched people's spiritual life, and we can easily obtain massive amounts of digital photos. However, due to the lack of professional guidance and differences in aesthetic appreciation, the photos taken many photographers lack aesthetics. This article is dedicated to the research of image aesthetics, using computers to simulate human perception, and realize the evaluation or beautification of images in line with human aesthetics. In terms of image classification, this article examines the unique perception of human vision on images and proposes new aesthetic features. Combining visual features and semantic features, the SVM algorithm is utilized to build an aesthetic classifier. In the aspect of image optimization, this paper uses the detection of the main image area and the division line of the area and adjusts the main body size and position of the image according to common aesthetic rules, so as to realize the optimization adjustment of the composition of the social image. The experimental results show that the accuracy of social image classification is 97.7%, and the optimized and adjusted images are more aesthetic. [ABSTRACT FROM AUTHOR]

Micro-expression is a special kind of human emotion. Due to its characteristics of short time, low intensity, and local region, micro-expression recognition is a difficult task. At the same time, it is a natural, spontaneous, and unconcealable emotion that can well convey a person's actual psychological state and, therefore, has certain research value and practical significance. This paper focuses on micro-expression recognition in the field of deep learning through the survey and understanding of existing micro-expression recognition research, as well as grasping the research trend, for the previous literature on micro-expression review ignored the handcrafted features as an important part of the micro-expression recognition framework, and at the same time lacked the analysis of the various enhancement processing, a new micro-expression recognition framework based on deep learning is proposed. The model is designed from the perspective of modularity and streaming data. On the other hand, unlike the previous process of feeding the data directly into the network for training and recognition, the handcrafted features are used as the initial encoding of the micro-expression recognition data, followed by the training and learning of the deep model and at the same time the modular embedding approach is used to incorporate the feature enhancement module, and finally the classification and recognition. The article provides a detailed summary and analysis of each part of the whole framework and a comprehensive introduction to the current problems, experimental protocols, evaluation metrics, and application areas. Finally, it summarizes and gives possible future research directions. Therefore, this paper provides a comprehensive summary and analysis of micro-expression recognition in deep learning so that the related personnel can have a new understanding of the development of this field. On the other hand, it proposes a new recognition framework that also provides a reference for the researchers' later research. [ABSTRACT FROM AUTHOR]

In order to improve the diagnosis and classification effect of pituitary tumors, this paper combines the current common machine learning methods and classification prediction methods to improve the traditional machine learning algorithms. Moreover, this paper analyzes the feature algorithm based on the feature extraction requirements of pituitary tumor pictures and compares a variety of commonly used algorithms to select a classification algorithm suitable for the model of this paper through comparison methods. In addition, this paper carries out the calculation of the prediction algorithm and verifies the algorithm according to the actual situation. Finally, based on the neural network algorithm, this paper designs and constructs the algorithm function module and combines the actual needs of pituitary tumors to build the model and verify the performance of the model. The research results show that the model system constructed in this paper meets the expected demand. [ABSTRACT FROM AUTHOR]

Natural disasters can be unpredictable and catastrophic. Even after the event, the repercussions are prolonged due to the incompetence of disaster management strategies. To mitigate the effects of a natural hazard, disaster management teams have to rapidly come forth with innovative plans of action. To contain the damage that a disaster causes, the response time, and preparedness of a disaster management team is crucial. Having a forewarning about the nature of the disaster can prove to be beneficial for the management team. Remote/unreachable areas such as deep parts of the forests, far-flung rural areas, oceans and other hard to reach locations are at a higher risk of receiving poorer aid and response due to the lack of communication and connectivity with the rest of the world. Our paper provides a solution for this particular problem by suggesting UAV/Drones for surveillance and monitoring in these inaccessible, disaster struck areas. The drones not only monitor the situation but constantly click pictures and send them to a base station, where the images are used to collect insights about the type of disaster that has to be dealt with. The paper proposes a deep learning model based on feature concatenation for classification of disasters which can be deployed at the base station, and can receive data in the form of images from a UAV/drone hovering over the affected place. The proposed model is efficient and able to achieve a higher accuracy as compared to the leading CNN models and closely related recent works as well. [ABSTRACT FROM AUTHOR]

Hand gestures are becoming an important part of the communication method between humans and machines in the era of fast-paced urbanization. This paper introduces a new standard dataset for hand gesture recognition, Static HAnd PosturE (SHAPE), with adequate side, variation, and practicality. Compared with the previous datasets, our dataset has more classes, subjects, or scenes than other datasets. In addition, the SHAPE dataset is also one of the first datasets to focus on Asian subjects with Asian hand gestures. The SHAPE dataset contains more than 34,000 images collected from 20 distinct subjects with different clothes and backgrounds. A recognition architecture is also presented to investigate the proposed dataset. The architecture consists of two phases that are the hand detection phase for preprocessing and the classification phase by customized state-of-the-art deep neural network models. This paper investigates not only the high accuracy, but also the lightweight hand gesture recognition models that are suitable for resource-constrained devices such as portable edge devices. The promising application of this study is to create a human–machine interface that solves the problem of insufficient space for a keyboard or a mouse in small devices. Our experiments showed that the proposed architecture could obtain high accuracy with the self-built dataset. Details of our dataset can be seen online at https://users.soict.hust.edu.vn/linhdt/dataset/ [ABSTRACT FROM AUTHOR]

With the increased use of social network sites, such as Twitter, attackers exploit these platforms to spread counterfeit content. Such content can be fake advertisements or illegal content. Classifying such content is a challenging task, especially in Arabic. The Arabic language has a complex structure and makes classification tasks more difficult. This paper presents an approach to classifying Arabic tweets using classical machine learning (non-deep machine learning) and deep learning techniques. Tweets corpus were collected through Twitter API and labelled manually to get a reliable dataset. For an efficient classifier, feature extraction is applied to the corpus dataset. Then, two learning techniques are used for each feature extraction technique on the created dataset using N-gram models (uni-gram, bi-gram, and char-gram). The applied classical machine learning algorithms are support vector machines, neural networks, logistics regression, and naïve Bayes. Global vector (GloVe) and fastText learning models are utilised for the deep learning approaches. The Precision, Recall, and F1-score are the suggested performance measures calculated in this paper. Afterwards, the dataset is increased using the synthetic minority oversampling technique class to create a balanced dataset. After applying the classical machine learning models, the experimental results show that the neural network algorithm outperforms the other algorithms. Moreover, the GloVe outperforms the fastText model for the deep learning approach. [ABSTRACT FROM AUTHOR]

In medical practice, all decisions, as for example the diagnosis based on the classification of images, must be made reliably and effectively. The possibility of having automatic tools helping doctors in performing these important decisions is highly welcome. Artificial Intelligence techniques, and in particular Deep Learning methods, have proven very effective on these tasks, with excellent performance in terms of classification accuracy. The problem with such methods is that they represent black boxes, so they do not provide users with an explanation of the reasons for their decisions. Confidence from medical experts in clinical decisions can increase if they receive from Artificial Intelligence tools interpretable output under the form of, e.g., explanations in natural language or visualized information. This way, the system outcome can be critically assessed by them, and they can evaluate the trustworthiness of the results. In this paper, we propose a new general-purpose method that relies on interpretability ideas. The approach is based on two successive steps, the former being a filtering scheme typically used in Content-Based Image Retrieval, whereas the latter is an evolutionary algorithm able to classify and, at the same time, automatically extract explicit knowledge under the form of a set of IF-THEN rules. This approach is tested on a set of chest X-ray images aiming at assessing the presence of COVID-19. [ABSTRACT FROM AUTHOR]

Feature extraction plays an important role in arrhythmia classification, and successful arrhythmia classification generally depends on ECG feature extraction. This paper proposed a feature extraction method combining traditional approaches and 1D-CNN aiming to find the optimal feature set to improve the accuracy of arrhythmia classification. The proposed method is verified by using the MIT-BIH arrhythmia benchmark database. It is found that the features extracted by 1D-CNN and discrete wavelet transform form the optimal feature set with the average classification accuracy up to 98.35%, which is better than the latest methods. [ABSTRACT FROM AUTHOR]

Automatic detection of falls is extremely important, especially in the remote monitoring of elderly people, and will become more and more critical in the future, given the constant increase in the number of older adults. Within this framework, this paper deals with the task of evaluating several artificial intelligence techniques to automatically distinguish between different activities of daily living (ADLs) and different types of falls. To do this, UniMiB SHAR, a publicly available data set containing instances of nine different ADLs and of eight kinds of falls, is considered. We take into account five different classes of classification algorithms, namely tree-based, discriminant-based, support vector machines, K-nearest neighbors, and ensemble mechanisms, and we consider several representatives for each of these classes. These are all the classes contained in the Classification Learner app contained in MATLAB, which serves as the computational basis for our experiments. As a result, we apply 22 different classification algorithms coming from artificial intelligence under a fivefold cross-validation learning strategy, with the aim to individuate which the most suitable is for this data set. The numerical results show that the algorithm with the highest classification accuracy is the ensemble based on subspace as the ensemble method and on KNN as learner type. This shows an accuracy equal to 86.0%. Its results are better than those in the other papers in the literature that face this specific 17-class problem. [ABSTRACT FROM AUTHOR]

Cross-domain few-shot classification aims to recognize images in the new categories and domains that only contain few but unacquainted images. Considering the problems of fine-grained recognition in cross-domain few-shot classification including marginal overall-discrepancy in feature distribution and obvious fine-grained difference in dataset, this paper proposes a simple and effective attentive fine-grained recognition (AFGR) model. Specifically, the residual attention module is stacked into the feature encoder based on the residual network, which can linearly enhance different semantic feature information to help the metric function better locate the fine-grained feature information of the image. In addition, a bilinear metric function structure is proposed to learn and fuse different fine-grained image features, respectively, since the weights of bilinear measurement functions are not shared. Eventually, the final classification result is obtained by merging the recognition of bilinear metric function through posterior probability multiplication. In this paper, ablation experiments and comparative experiments are carried out with the typical few-shot dataset mini-ImageNet as the training domain and the CUB, Cars, Places and Plantae datasets as the test domain. The experimental results demonstrate that the proposed AFGR method is effective, with the highest increase in recognition accuracy 13.82% and 7.95% compared with the latest results under the experimental settings of 5-way1-shot and 5-way5-shot, respectively, which also proves the problems of fine-grained recognition in cross-domain small sample classification. [ABSTRACT FROM AUTHOR]

The amount of digital data in the universe is growing at an exponential rate with the rapid development of digital information, and this reveals new machine learning methods. Learning algorithms using hyperboxes are a subsection of machine learning methods. Fuzzy min–max neural network (FMNN) are one of the most common and advanced methods using hyperboxes. FMNN is a special type of NeuroFuzzy system that combines the artificial neural network and fuzzy set into a common framework. This paper conducts an extensive bibliometric and network analysis of FMNN literature. Two hundred and sixty-two publications are analysed from the period of 1992–2022. Several analyses are realized in order to identify trends, challenges and key points in a more scientific and objective way that affect the development of knowledge in the FMNN domain. It can be seen from bibliometric analysis that there is rapid development in the last 10 years. Social network analysis results show that Chee Peng Lim is the most active author in the network. Besides, the modifications of FMNN are generally developed for classification. However, there are still potential future research opportunities for clustering. [ABSTRACT FROM AUTHOR]

Deep learning techniques have gained immense popularity recently because of their remarkable capacity to learn complex patterns and features from large datasets. These techniques have revolutionized many fields by achieving advanced performance in various tasks. The availability of large datasets and the advancement of computing resources have enabled deep learning models to perform well in solving challenging problems. As a result, they have become an essential tool in many industries, including agriculture. The application of deep learning in agriculture has great potential for increasing productivity, reducing costs, and improving sustainability by aiding in the early identification and prevention of plant leaf diseases, optimizing crop yields, and facilitating precision agriculture. This paper suggests using a novel approach to automatically classify multi-class leaf diseases in tomatoes using a deep multi-scale convolutional neural network (DMCNN). The proposed DMCNN architecture consists of parallel streams of convolutional neural networks at different scales, which get merged at the end to form a single output. The images of tomato leaves are preprocessed using data augmentation techniques and fed into the DMCNN model to classify disease. The proposed approach is evaluated on a dataset of tomato plant images containing 10 distinct classes of diseases and compared with different existing models. The research results reveal that the suggested DMCNN model performs better than other models in terms of accuracy, precision, recall, and F1 score. Furthermore, the proposed model reported an overall accuracy of 99.1%, which is higher than the accuracy of existing models tested on the same dataset. The study demonstrates the potential of deep learning techniques for automated disease classification in agriculture, which can aid in early disease detection and prevent crop loss. [ABSTRACT FROM AUTHOR]

This paper presents a methodology for developing a volcano-seismic event classification system using a multi-station deep learning approach to support monitoring the Nevados del Chillán Volcanic Complex, which has been active since 2017. A convolutional network of multiple inputs processes the information from an event recorded up to five seismic stations. Each record is represented by its normalized spectrogram; thus, the network may receive from one to five spectrograms as input. The design includes entering additional information into the network, like the stations configuration and the event duration, information not provided by the spectrograms. Finally, this work includes the design and implementation of a relational database to access the continuous traces of events, showing different subsets of data quickly and efficiently. The results show that the classification of an event recorded up to five stations is substantially more effective than a single-station strategy. However, incorporating additional information of the signal does not significantly improve the classification performance. [ABSTRACT FROM AUTHOR]

Twitter produces a massive amount of data due to its popularity that is one of the reasons underlying big data problems. One of those problems is the classification of tweets due to use of sophisticated and complex language, which makes the current tools insufficient. We present our framework HTwitt, built on top of the Hadoop ecosystem, which consists of a MapReduce algorithm and a set of machine learning techniques embedded within a big data analytics platform to efficiently address the following problems: (1) traditional data processing techniques are inadequate to handle big data; (2) data preprocessing needs substantial manual effort; (3) domain knowledge is required before the classification; (4) semantic explanation is ignored. In this work, these challenges are overcome by using different algorithms combined with a Naïve Bayes classifier to ensure reliability and highly precise recommendations in virtualization and cloud environments. These features make HTwitt different from others in terms of having an effective and practical design for text classification in big data analytics. The main contribution of the paper is to propose a framework for building landslide early warning systems by pinpointing useful tweets and visualizing them along with the processed information. We demonstrate the results of the experiments which quantify the levels of overfitting in the training stage of the model using different sizes of real-world datasets in machine learning phases. Our results demonstrate that the proposed system provides high-quality results with a score of nearly 95% and meets the requirement of a Hadoop-based classification system. [ABSTRACT FROM AUTHOR]

One of the most challenging issues in pattern recognition is the data attribution selection process. Feature selection plays a key role in solving problems with high-dimensional data and is a fundamental step in pre-processing many classifications and machine learning problems. The feature selection method reduces the amount of data and increases the category precision. Unrelated data, which can lead to inappropriate classification, is thus removed to obtain fewer features. In this paper, the Binary Gray Wolf Optimization algorithm uses the Wrapper method for feature selection. The transfer function is an essential part of BGWO to map a continuous value to a binary value. In this study, eight transfer functions are divided into two families: S-shaped and V-shaped. Previous research has used only one transfer function for the whole algorithm, and all wolves in the whole algorithm deal with this transfer function. In this paper, each wolf has its own transfer function. Because algorithms are evolutionary meta-innovations and can optimize themselves, each wolf can play a role in the whole algorithm at any stage while optimizing itself and adapting to its community, and not just depend on one transfer function. In the proposed method, eight transfer functions are divided into two families, S-shaped and V-shaped. This article proposes two approaches for learning the transfer function, by selecting the transfer function and the slope of these functions. In the first approach, we add three or two binary bits to the initial population. If two bits are added, four modes of the transfer function are available, and if three bits are added, eight transfer functions are achievable. These bits are used as a criterion for selecting a predefined transfer function for each wolf. So, in the proposed method, each wolf has its transfer function. During the implementation of the algorithm, the wolves update their position according to the evaluation function and learning. In the second approach, ten or twenty-one binary bits are added to the initial population. If ten binary bits are used, we will have a transfer function, and 2 10 coefficient modes are available for the slope of the transfer function. If twenty-one binary bits are used, we have two transfer functions available. So, there are 2 10 modes for the gradient of the transfer function. These bits are used as a criterion for selecting the transfer function and the coefficient affecting the slope of these functions. In both ideas, after each iteration of the algorithm, the position of the wolves is updated and based on the evaluation function, the alpha wolf is identified and the transfer function is selected. With subsequent iterations, the algorithm learns and optimizes the transfer function to achieve the best feature selection with the smallest error. Experimental results on ten UCI datasets show that selecting the obtained feature subset with high classification accuracy is efficient. [ABSTRACT FROM AUTHOR]

Few-shot learning is to learn to discriminate novel classes from a minimal amount of support images. The core of the matter lies in obtaining effective feature representations from limited samples and measuring the similarity between query and support images. In this paper, we approach this problem from two perspectives: effective feature learning and similarity measurement. We observe that the objects being classified are typically the common salient targets within the support class in few-shot classification. Based on this observation, we design a Co-Salient Feature Extraction module which can learn the correlations among the intra-class samples in the support images, thereby making the learned features more representative of the class. Regarding similarity measurement, we comprehensively consider both the global features and local details of the images and propose a Multi-Scale Metric module to implement holistic metric and improve the reliability of image-to-class measurement. We conduct experiments on four benchmark datasets for few-shot image classification, including general object recognition, fine-grained categorization, and cross-domain classification. The experimental results in various few-shot learning scenarios demonstrate that the proposed Intra-class Co-Salient (ICoS) network achieves competitive performance, particularly excelling in fine-grained classification, ICoS outperforms the similarity techniques DN4 by 20.61% in 1-shot and 11.23% in 5-shot on the CUB dataset, demonstrating the validity of co-salient learning. [ABSTRACT FROM AUTHOR]

Generalized zero-shot learning (GZSL) aims to identify the samples that belong to a different label space than training samples, while the samples of these unseen classes have the same semantic space as the samples of seen classes. Therefore, most GZSL models use the samples to train the fundamental part, then use semantic information to match or generate the features of the unseen classes as the generalized part. However, the importance of features and their consistency with semantics are different. This difference causes jeopardizing between each other, becoming one of the main reasons that the classification results of seen and unseen classes in the GZSL are lower than the traditional classification and the CZSL. To maintain the training effect from two parts without jeopardizing each other, IDN-GZSL is presented in this paper. We provide rearrangement modules in different isolated scenarios to distinguish the positive and negative parts of the feature and extract the positive features. Three rearrangement modules have been trained using existing training data and generated data in corresponding scenarios. Although the three scenarios are isolated, the decoupling and relation modules are used to support the rearrangement and acquire positive features for semantic information, and the distillation module is used to establish indirect connections. The results demonstrate that our methods take effect with three significant improvements in ablation experiments: successfully trained rearrangement modules for obtaining positive features in multiple scenarios for classification, eliminated jeopardizing among different scenarios by establishing three isolation scenarios, and established necessary connections among them by distillation module. Our method performs the best classification ability of unseen classes on four datasets and achieves the best performance on AWA2, CUB, and FLO when evaluating the method through harmonic means. [ABSTRACT FROM AUTHOR]

With the rapid development of the Internet, image information is explosively growing. Traditional image classification methods are difficult to deal with huge image data and cannot meet people's requirements on the accuracy and speed of image classification. In recent years, the convolutional neural network (CNN) has been developing rapidly, and it has performed extremely well. The image classification method based on CNN breaks through the bottleneck of traditional image classification methods and becomes the mainstream image classification algorithm at present. CT image classification algorithm is one of the research hot spots in the field of medical image. The purpose of this paper is to apply convolutional neural network to CT image classification, so as to speed up CT image classification and improve the accuracy of CT image classification and so as to reduce the workload of doctors and improve work efficiency. In this paper, CT images are classified by CDBN model. Vector machine SVM is used as the feature classifier of CDBN model to enhance feature transfer and reuse so as to enrich the features. It also suppresses features that are not very useful for current tasks and improves the performance of the model. Using CDBN to classify CT images, several commonly used gray images are compared. Comparing the results of the ordinary gradient algorithm with Adam algorithm, we can get the CDBN model using Adam optimization algorithm. In CT image classification, both accuracy and speed have a good effect. The experimental results show that the training speed of CDBN model of Adam optimization algorithm in CT image classification is 3% faster than that of general gradient algorithm. [ABSTRACT FROM AUTHOR]

The monitoring of physical activities and recognition of motion disorders belong to important diagnostical tools in neurology and rehabilitation. The goal of the present paper is in the contribution to this topic by (1) analysis of accelerometric signals recorded by wearable sensors located at specific body positions and by (2) implementation of deep learning methods to classify signal features. This paper uses the general methodology to analysis of accelerometric signals acquired during cycling at different routes followed by the global positioning system. The experimental dataset includes 850 observations that were recorded by a mobile device in the spine area (L3 verterbra) for cycling routes with the different slope. The proposed methodology includes the use of deep learning convolutional neural networks with five layers applied to signal values transformed into the frequency domain without specification of any signal features. The accuracy of discrimination between different motion patterns for the uphill and downhill cycling and recognition of 4 classes associated with different route slopes was 96.6% with the loss criterion of 0.275 for sigmoidal activation functions. These results were compared with those evaluated for selected sets of features estimated for each observation and classified by the support vector machine, Bayesian methods, and the two-layer neural network. The best cross-validation error of 0.361 was achieved for the two-layer neural network model with the sigmoidal and softmax transfer functions. Our methodology suggests that deep learning neural networks are efficient in the assessment of motion activities for automated data processing and have a wide range of applications, including rehabilitation, early diagnosis of neurological problems, and possible use in engineering as well. [ABSTRACT FROM AUTHOR]

This paper considers a category of classification problems where the samples of different classes are not represented equally. It arises in a variety of application areas and has been widely studied in pattern recognition. This paper focuses on enhancing the original data representation by combining the gravitation-based method with multiple empirical kernel approach, this paper proposes a sample level method known as the gravitational balanced multiple kernel learning (GBMKL) method. Our proposed GBMKL method integrates gravity strategy to generate the gravitation balanced midpoint samples (GBMS) located on the classification boundary; meanwhile, the classification boundary can be rectified by the nearest neighbors of the boundary (NNB) samples, which can improve the generalization performance. We further design two regularization terms corresponding to GBMS and NNB to avoid overfitting. In the training and testing process, the samples are mapped into multiple empirical kernel spaces to obtain more sufficient data representation. We conduct extensive computational experiments on 54 imbalanced datasets including both artificial and real-word datasets selected from knowledge extraction based on evolutionary learning datasets, the experimental results reveal interesting insights and show the advantages of the proposed GBMKL approach for dealing with the imbalanced classification problems. In addition, parameter analysis of two regularization terms confirms their positive impacts on the classification performance. [ABSTRACT FROM AUTHOR]

With the ever-improving degree of car privatization, the number of car-owning residents has repeatedly reached new highs, but it has also caused traffic congestion, especially lane congestion. To reduce emergency lane congestion, and realize the intelligent detection and classification of emergency lanes, this paper introduces logistic regression theory and proposes a vehicle detection method based on logistic regression. Based on the powerful feature abstraction and automatic learning capabilities of convolutional networks, it studies vehicles based on convolutional networks. The recognition algorithm analyses the contribution of the convolution kernel in the convolutional network, and transforms the original emergency lane vehicle detection. The research results show that the emergency lane vehicle detection rate based on logistic regression and a deep convolutional network designed in this paper can reach approximately 98%, which is higher than the detection methods based on SVM and AdaBoost, and the detection time is not different from SVM and AdaBoost. After training, the convolution kernel of the deep convolutional network, the detection time and detection accuracy improved to a certain extent. This shows that the emergency lane vehicle detection and classification method based on logistic regression and a deep convolutional network, can play an important role in emergency lane intelligence. [ABSTRACT FROM AUTHOR]

The paper is a comprehensive review of the current research trends in the area of Arabic language especially state-of-the-art approaches to highlight the current status of diverse research aspects of that area to facilitate the adaption and extension of previous systems into new applications and systems. The Arabic language has deep, widespread and unexplored scope to research although the tremendous effort and researches that had been done previously. Modern state-of-the-art methods and approaches with fewer errors are required according to the high speed of hardware and technology development. The focus of this article will be on the offline Arabic handwritten text recognition as it is one of the most important topics in the Arabic scope. The main objective of this paper is critically analyzing the current researches to identify the problem areas and challenges faced by the previous researchers. This identification is intended to provide many recommendations for future advances in the area. It also compares and contrasts technical challenges, methods and the performances of handwritten text recognition previous researches works. It summarizes the critical problems and enumerates issues that should be considered when addressing these tasks. It also shows some of the Arabic datasets that can be used as inputs and benchmarks for training, testing and comparisons. Finally, it provides a fundamental comparison and discussion of some of the remaining open problems and trends in that field. [ABSTRACT FROM AUTHOR]

Skin cancer, primarily resulting from the abnormal growth of skin cells, is among the most common cancer types. In recent decades, the incidence of skin cancer cases worldwide has risen significantly (one in every three newly diagnosed cancer cases is a skin cancer). Such an increase can be attributed to changes in our social and lifestyle habits coupled with devastating man-made alterations to the global ecosystem. Despite such a notable increase, diagnosis of skin cancer is still challenging, which becomes critical as its early detection is crucial for increasing the overall survival rate. This calls for advancements of innovative computer-aided systems to assist medical experts with their decision making. In this context, there has been a recent surge of interest in machine learning (ML), in particular, deep neural networks (DNNs), to provide complementary assistance to expert physicians. While DNNs have a high processing capacity far beyond that of human experts, their outputs are deterministic, i.e., providing estimates without prediction confidence. Therefore, it is of paramount importance to develop DNNs with uncertainty-awareness to provide confidence in their predictions. Monte Carlo dropout (MCD) is vastly used for uncertainty quantification; however, MCD suffers from overconfidence and being miss calibrated. In this paper, we use MCD algorithm to develop an uncertainty-aware DNN that assigns high predictive entropy to erroneous predictions and enable the model to optimize the hyper-parameters during training, which leads to more accurate uncertainty quantification. We use two synthetic (two moons and blobs) and a real dataset (skin cancer) to validate our algorithm. Our experiments on these datasets prove effectiveness of our approach in quantifying reliable uncertainty. Our method achieved 85.65 ± 0.18 prediction accuracy, 83.03 ± 0.25 uncertainty accuracy, and 1.93 ± 0.3 expected calibration error outperforming vanilla MCD and MCD with loss enhanced based on predicted entropy. [ABSTRACT FROM AUTHOR]

The selection of proper classifiers for a given data set is full of challenges. The critical problem of classifier selection is how to extract feature from data sets. This paper proposes a new method for feature extraction of a data set. Our method not only preserves the geometrical structure of a data set, but also characterizes the decision boundary of classification problems. Specifically speaking, the extracted feature can recover a data set that has the same Euclidean geometrical structure as the original data set. We present an efficient algorithm to compute the similarity between data set features. We theoretically analyze how the similarity between our features affects the performance of the support vector machine, a well-known classifier. The empirical results show that our method is effective in finding suitable classifiers. [ABSTRACT FROM AUTHOR]

To reliably and accurately classify complicated "big" datasets, machine learning models must be continually improved. This research proposes straightforward yet competitive neural networks for text classification, even though graph neural networks (GNN) have reignited interest in graph-based text classification models. Convolutional neural networks (CNN), artificial neural networks (ANN), and their refined "fine-tuned" models (denoted as FT-CNN and FT-ANN) are the names given to our proposed models. The models presented in this paper demonstrate that our simple models like (CNN, ANN, FT-CNN, and FT-ANN) can perform better than more complex GNN ones such as (SGC, SSGC, and TextGCN) and are comparable to others (i.e., HyperGAT and Bert). The process of fine-tuning is also highly recommended because it improves the performance and reliability of models. The performance of our suggested models on five benchmark datasets (namely, Reuters (R8), R52, 20NewsGroup, Ohsumed, and Mr) is vividly illustrated. According to the experimental findings, on the majority of the target datasets, these models—especially those that have been fine-tuned—perform surprisingly better than SOTA approaches, including GNN-based models. [ABSTRACT FROM AUTHOR]

Big data is a large set of data that is analyzed with the calculation to manifest myriad sources. Big data is capable of handling various challenges to processing huge amounts of data. To handle issues based on large-scale databases, a MapReduce framework is employed which provides robust and simple infrastructure for huge datasets. This paper proposes a novel Elastic collision seeker optimization based Faster R-CNN (ECSO-FRCNN) classifier for efficient big data classification. The proposed ECSO-FRCNN classifier is capable of handling missing attributes, and incremental learning and improves training performance effectively. As the proposed technique deals with large data samples, it necessitates the inclusion of the MapReduce framework. The adaption of MapReduce design in big data classification prevents the classification results from uncertainties such as data redundancy, misclassification, and storage issues. The proposed method is examined with three standard datasets, namely the skin segmentation dataset, mushroom dataset, and localization dataset, collected from the University of California, UCI machine learning repository. Finally, extensive experimental analysis is carried out for various parameters to depict the efficiency of the system. [ABSTRACT FROM AUTHOR]

This paper investigates a wireless blockchain network with mobile edge computing in which Internet of Things (IoT) devices can behave as blockchain users (BUs). This blockchain network's ultimate goal is to increase the overall profits of all BUs. Because not all BUs join in the mining process, using traditional swarm and evolution algorithms to solve this problem results in a high level of redundancy in the search space. To solve this problem, a modified chaotic Henry single gas solubility optimization algorithm, called CHSGSO, has been proposed. In CHSGSO, the allocation of resources to BUs who decide to engage in mining as an individual is encoded. This results in a different size for each individual in the entire population, which leads to the elimination of unnecessary search space regions. Because the individual size equals the number of participating BUs, we devise an adaptive strategy to fine-tune each individual size. In addition, a chaotic map was incorporated into the original Henry gas solubility optimization to improve resource allocation and accelerate the convergence rate. Extensive experiments on a set of instances were carried out to validate the superiority of the proposed CHSGSO. Its efficiency is demonstrated by comparing it to four well-known meta-heuristic algorithms. [ABSTRACT FROM AUTHOR]

Micro-ribonucleic acids (miRNAs) are tiny noncoding ribonucleic acid (RNA) molecules that involve various biological processes for cancer advancements. Classification of cancer is a crucial problem in microarray technology because of the huge number of miRNAs and a limited number of samples. Therefore, to improve the classification of cancer between benign and malignant samples, this paper presents a hybrid filter-wrapper method that uses the multi-filter ensemble (MFE) approach as a filter method used in selecting the top-ranked miRNAs. For the wrapper approach, three different algorithms based on simultaneous utilization of the grey wolf optimizer (GWO), evolutionary population dynamics (EPD), and two selection mechanisms, are presented to enhance the GWO algorithm performance. The first approach involves the application of the EPD mechanisms to the GWO to eliminate the worst solutions of GWO and repositions them toward alpha, beta, or delta wolves to improve exploitation. So also, the GWO needed to randomly reinitialize its poor solutions near to the search space, using the EPD for the enhancement of the exploration process. In the second and third approaches, the Roulette wheel selection (RWS) and tournament selection (TS) are applied to offer a chance for low fitness individuals to be selected during the search process that maintains the diversity of the selected solutions. The proposed GWO_EPD algorithms are applied for miRNAs selection, and obtained results demonstrate that the EPD can improve the search capability of GWO in terms of avoiding local optima, exploration, and convergence rate. Moreover, the results of GWO_EPD are compared with twelve miRNAs subset selection approaches using classification accuracy and the number of selected miRNAs. [ABSTRACT FROM AUTHOR]

The spread of the COVID-19 started back in 2019; and so far, more than 4 million people around the world have lost their lives to this deadly virus and its variants. In view of the high transmissibility of the Corona virus, which has turned this disease into a global pandemic, artificial intelligence can be employed as an effective tool for an earlier detection and treatment of this illness. In this review paper, we evaluate the performance of the deep learning models in processing the X-Ray and CT-Scan images of the Corona patients' lungs and describe the changes made to these models in order to enhance their Corona detection accuracy. To this end, we introduce the famous deep learning models such as VGGNet, GoogleNet and ResNet and after reviewing the research works in which these models have been used for the detection of COVID-19, we compare the performances of the newer models such as DenseNet, CapsNet, MobileNet and EfficientNet. We then present the deep learning techniques of GAN, transfer learning, and data augmentation and examine the statistics of using these techniques. Here, we also describe the datasets introduced since the onset of the COVID-19. These datasets contain the lung images of Corona patients, healthy individuals, and the patients with non-Corona pulmonary diseases. Lastly, we elaborate on the existing challenges in the use of artificial intelligence for COVID-19 detection and the prospective trends of using this method in similar situations and conditions. [ABSTRACT FROM AUTHOR]

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]

This paper presents a novel, fully automatic method to grayscale images colorization using a scene guided neural network. In our proposed method, given a training set of both grayscale images and their corresponding color images, we first extract features of each grayscale pixel. These features, together with the corresponding RGB values of that pixel are input to train a colorization neural network for each given scene. To improve the performance of colorization, in both speed and results, we further classify the input and training images into different scene classes. We adopt a linear image classification method to generate a scene guided codebook and use it to determine the scene class of the input image. The preliminary colorization result is then generated by the corresponding trained neural network of the scene class of the input image. Finally, an image guided filter is used to refine colorized images. Inspired by the recent success in deep learning techniques which provide stabilizing modeling of large-scale medical image data, the proposed paper formulating the enhancement and colorization problem, so that colorization techniques can be directly used to ensure medical images with high quality. The experimental results on a broad range of images demonstrate that our method has better colorization performance as compared to that of the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

To date, Artificial Intelligence systems for handwriting and drawing analysis have primarily targeted domains such as writer identification and sketch recognition. Conversely, the automatic characterization of graphomotor patterns as biomarkers of brain health is a relatively less explored research area. Despite its importance, the work done in this direction is limited and sporadic. This paper aims to provide a survey of related work to provide guidance to novice researchers and highlight relevant study contributions. The literature has been grouped into "visual analysis techniques" and "procedural analysis techniques". Visual analysis techniques evaluate offline samples of a graphomotor response after completion. On the other hand, procedural analysis techniques focus on the dynamic processes involved in producing a graphomotor reaction. Since the primary goal of both families of strategies is to represent domain knowledge effectively, the paper also outlines the commonly employed handwriting representation and estimation methods presented in the literature and discusses their strengths and weaknesses. It also highlights the limitations of existing processes and the challenges commonly faced when designing such systems. High-level directions for further research conclude the paper. [ABSTRACT FROM AUTHOR]

Multiple sclerosis (MS) is one of the most common inflammatory neurological diseases in young adults. There are three types of MS: (1) In relapsing remitting MS (RRMS), people have temporarily periods of relapses (attacks) for days or weeks, and then symptoms seem to disappear (remitting stage). (2) In secondary progressive MS (SPMS), symptoms worsen more steadily over time. Attacks (relapses) may occur time to time but the disease can progress in non-attack periods. It is estimated that half of the RRMS patients progress to SPMS in 10 years. (3) Primary progressive MS (PPMS) is characterized by slowly worsening symptoms from the beginning, with no relapses or remissions. For PPMS patients, disability progresses slowly. Researchers have found out that in the first year, MS causes more damage than following 5–10 years. Therefore, early diagnosis is vital. In this context, deep learning models started to be popular for assisting identification/diagnosis/classification of MS patients using magnetic resonance imaging (MRI). This paper provides an in-depth review of deep learning approaches for identification and classification of MS using brain MRI images. We discuss recent trends of deep learning methods for MS identification under three categories: CNN models, hybrid models (CNN with a classifier) and deep transfer learning models. Existing deep learning algorithms are analyzed and compared according to their architecture, image modality, pre-processing, feature extraction, classifier, dataset, categories and accuracy. This survey paper would provide a valuable source for researchers who are interested in state-of-the-art deep learning methods for MS identification using MRI images. [ABSTRACT FROM AUTHOR]

Universum refers to additional samples which contain priori knowledge for classification but belonging to none of the class. It has been proved that universum positioned “in between” the two classes obtain better results. Since opinions on stock market defined as investor sentiment involve quite a number of neutral views, these neutral views can be used as universum samples to better identify investor sentiment. With universum samples, this paper uses support vector machine (SVM) to classify posts on stock forum. We define bullish views as positive samples, define bearish views as negative samples, and also further discuss the situation of a 3-class problem with neutral views. Compared with standard SVM, the empirical studies with universum samples in this paper show better performance for both 2- and 3-class classifications. [ABSTRACT FROM AUTHOR]

Traditional methods for unsupervised cross-domain classification learn a common low-dimensional subspace using images from a well-labeled source domain and an unlabeled target domain. To achieve domain-invariant feature extraction, researchers typically focus on minimizing the distribution discrepancy. However, these methods often overlook the fact that label information contains critical categorization information for discriminative subspace learning. This paper proposes a novel method, named domain-invariant feature learning with label information integration (DILI), which integrates metric learning and label information extraction to learn a cross-domain discriminant subspace. DILI first reduces the distances between the source and target domains to mitigate the marginal distribution discrepancy. Then, it reduces the distances between cross-domain samples from the same class to mitigate the conditional distribution discrepancy. Dual terms are imposed to balance the label information of both domains to learn common features, and a discriminant subspace is learned for cross-domain tasks. This method obtains domain-invariant features, while the label information of the source domain and pseudo labels from the target domain are used to improve the discriminant of subspaces. Experimental results on eight cross-domain datasets show that DILI outperforms some state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Object detection has achieved great progress and is used in various tasks. However, detecting small objects with lack of appearance information is still a challenging task. It is found that even if the training set with rich images is used to train a network, the small objects in the image with different background complexity cannot be well detected. To address the above issue, this paper proposes a novel small object detection framework based on a background complexity classification strategy specific to the contradiction by adopting the idea of "divide and rule" in philosophy. Firstly, a Background Complexity Classification Network (BCCResNet) is proposed to coarsely classify the input images into three categories according to the complexity of their background textures. Then, a detection network with parallel structure is designed by using mainstream models to detect small objects for three categories of images. Extensive experiments are conducted on two small object detection datasets, i.e., AI-TOD and DOTAv1.0. Our proposed method can significantly improve AP of small object detection, showing effectiveness and advantages. [ABSTRACT FROM AUTHOR]

Emotion recognition using physiological signals has gained significant attention in recent years due to its potential applications in various domains, such as healthcare and entertainment. EEG signals have been particularly useful in emotion recognition due to their non-invasive nature and high temporal resolution. However, the development of accurate and efficient algorithms for emotion classification using EEG signals remains a challenging task. This paper proposes a novel hybrid algorithm for emotion classification based on EEG signals, which combines multiple adaptive network models and probabilistic neural networks. The research aims to improve the recognition accuracy of three and four emotions, which has been a challenge for existing approaches. The proposed model consists of N adaptively neuro-fuzzy inference system (ANFIS) classifiers designed in parallel, in which N is the number of emotion classes. The selected features with the most appropriate distribution for classification are given as input vectors to the ANFIS structures, and the system is trained. The outputs of these trained ANFIS models are combined to create a feature vector, which provides the inputs for adaptive networks, and the system is trained to acquire the emotional recognition output. The performance of the proposed model has been evaluated for classification on well-known emotion benchmark datasets, including DEAP and Feeling Emotions. The study results indicate that the model achieves an accuracy rate of 73.49% on the DEAP datasets and 95.97% on the Feeling Emotions datasets. These results demonstrate that the proposed model efficiently recognizes emotions and exhibits a promising classification performance. [ABSTRACT FROM AUTHOR]

In this paper, we study the post-hoc calibration of modern neural networks, a problem that has drawn a lot of attention in recent years. Despite the plethora of calibration methods proposed, there is no consensus yet on the inherent complexity of the task and, while some authors claim that simple functions solve the problem, others suggest that more expressive models are needed to capture misscalibration. As a first approach, we focus on the task of confidence scaling, specifically on post-hoc methods that generalize Temperature Scaling, which we refer to as the Adaptive Temperature Scaling family. We begin by demonstrating that while complex models like neural networks provide an advantage when there is ample data, they fail in scenarios where it is limited, notably common in fields like medical diagnosis. We then show how under this ideal data conditions the more expressive methods learn a relationship between the entropy of a prediction and its level of overconfidence, and based on this observation, we propose Entropy-based Temperature Scaling, a simple method that scales the confidence of a prediction according to this relationship. Results show that our method obtains state-of-the-art performance and is robust against data scarcity. Moreover, our proposed model enables a deeper understanding of the calibration process by the interpretation of the entropy as a measure of uncertainty in the network outputs. [ABSTRACT FROM AUTHOR]

Tailoring a deep convolutional neural network (DCNN) is a tedious and time-consuming task in the field of medical image analysis. In this research paper, Deep-generative adversial neural network (Deep-GAN) based model is proposed using grid search optimization (GSO) technique for identification and classification of thyroid nodule. The main objective of this work is to propose a deep learning (DL) model for the identification and classification of thyroid nodules without user or specialist intervention. The proposed model has gone through four phases namely (i) data acquisition, (ii) pre-processing (iii) data augmentation using GAN technique and (iv) optimization and classification using Deep-GAN model. Two pre-trained architectures namely Alex-Net and Visual Geometry Group (VGG-16) are considered for the identification and classification of thyroid nodule in ultrasonography (USG) images. From the experiment, it is found that Alex-GAN model has shown an improvement of 2 to 4 percentage points in comparison with VGG-GAN model and reported literature on Thyroid digital image database (TDID) public and collected dataset. [ABSTRACT FROM AUTHOR]

Graph convolutional networks-based text classification methods have shown impressive success in further improving the classification results by considering the structural relationship between words and texts. However, existing GCN-based text classification methods tend to ignore the semantic representation of the node and the global structural information among nodes. Besides, only the word granularity information within the text, i.e., endogenous source, is used to represent the text. Furthermore, the existing graph convolutional network approaches are faced with major challenges to handle large and dense graphs, i.e., neighbor explosion and noisy inputs. To address these shortcomings, this paper proposes an inductive learning-based text classification method that utilizes representation learning on heterogeneous information networks and exogenous knowledge. Firstly, a weighted heterogeneous information network for text (HINT) is constructed by introducing exogenous knowledge, in which the node types cover text, entities and words. The unstructured text is represented as a structured heterogeneous information network, which expands the granularity of text features and makes full use of the exogenous structural information and explicit semantic information to enhance the interpretability of text information. Besides, we also enhanced the graph neural network against the challenges of neighbor explosion and noisy inputs derived from HINT using two strategies: graph sampling and Dropedge, for semi-supervised learning with improved classification performance. The effectiveness of our model is demonstrated by examining four publicly available text classification datasets. Based on experimental results, our approach achieves state-of-the-art performance on the text classification datasets. [ABSTRACT FROM AUTHOR]

Enhanced machining quality, including the appropriate surface roughness of the machined parts, is the focus of many industries. This paper proposes and implements transformer-based deep learning (DL) architecture for machining roughness classification for the end-milling operation using cutting force and machining sound data. To increase the accuracy of the classification outcomes, audio feature extraction techniques—Mel-spectrogram and Mel-frequency cepstral coefficients (MFCCs)—were incorporated with the DL model. To measure and demonstrate the proposed model's performance, a number of experiments were conducted by training the models on 0–30 s machining data, including end mill-workpiece impact at the beginning of the experiment and 10–40 s machining data. Based on the outcomes' accuracies, four DL models were designed and the number of parameters, maximum epoch required for convergence, and training, validation and testing accuracies were compared for each model. DL models trained on 10–40 s machining data achieved over 90% validation and test accuracies, suggesting that the cutting force and machining data after reaching steady-state performs better than data from the beginning of machining operation. Confusion matrices were plotted for the inference results of each model to observe the prediction accuracy visually. The proposed method will be able to predict machining surface quality with high accuracy saving time and cost for post-operations. [ABSTRACT FROM AUTHOR]

Zero-shot learning (ZSL) aims at recognizing instances from unseen classes via training a classification model with only seen data. Most existing approaches easily suffer from the classification bias from unseen to seen categories since the models are only trained with seen data. In this paper, we tackle the task of ZSL with a novel Unseen Prototype Learning (UPL) model, which is a simple yet effective framework to learn visual prototypes for unseen categories from the corresponding class-level semantic information, and the learned features are treated as latent classifiers directly. Two types of constraints are proposed to improve the performance of the learned prototypes. Firstly, we utilize an autoencoder framework to learn visual prototypes from the semantic prototypes and reconstruct the original semantic information by a decoder to ensure the prototypes have a strong correlation with the corresponding categories. Secondly, we employ a triplet loss to make the average of visual features per class supervise the learned visual prototypes. In this way, the visual prototypes are more discriminative and as a result, the classification bias problem can be alleviated well. Besides, based on the episodic training paradigm in meta-learning, the model can accumulate wealthy experiences in predicting unseen classes. Extensive experiments on four datasets under both traditional ZSL and generalized ZSL show the effectiveness of our proposed UPL method. [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]

When it comes to offering song suggestions to a user, the music genre is one of the most used tags considered by music streaming services. Motivated by the growing number of songs available, automatic music genre classification systems have become a valuable tool for the creation of user-personalised playlists. Considering that feature engineering represents a major task to be addressed when one develops such systems, this work discusses the generation of new handcrafted features over songs, originally exploring high-order features' moments combined with their derivatives. Additionally, this paper proposes a new wrapper-based selection procedure rigorously based on statistical tests to identify a subset of features that maximise the performance of these systems, irrespective of the classification approach adopted, named Robust Selector of Basis Feature Sets. Based on a synergistic combination of both strategies, a compact subset with 81 features is derived over the GTZAN Dataset. When compared with alternative solutions, this feature set boosted the classification accuracy in datasets containing a wide range of music genres, such as ISMIR2004, BALLROOM, HOMBURG, and FMA Datasets. [ABSTRACT FROM AUTHOR]

The classification of environmental sounds is important for emerging applications such as automatic audio surveillance, audio forensics, and robot navigation. Existing techniques combined multiple features and stacked many CNN layers (very deep learning) to reach the desired accuracy. Instead of using many features and going deeper by stacking layers that are resource extensive, this paper proposes a novel technique that uses only a single feature, namely the Mel-Frequency Cepstral Coefficient (MFCC) and just three layers of CNN. We demonstrate that such a simple network can considerably outperform several conventional and deep learning-based algorithms. Through parameters fine-tuning of the data input, we reported a model that is significantly less complex in the architecture yet has recorded a similar accuracy of 95.59% compared to state-of-the-art deep models on UrbanSound8k dataset. [ABSTRACT FROM AUTHOR]

Holistic-based or segmentation-free handwritten word recognition is one of a pattern recognition problem that aims to recognize the entire word image as a single entity. It is a new modality that recognizes handwritten words from its overall shape and performs better than its complement, i.e., analytic approach for given small lexicon size. Due to technological advancements, society is becoming paperless and prefers to use digital platform for various tasks. This paper deals with the use of holistic approach for the recognition of offline handwritten Devanagari words based on some statistical features. Feature vector sets have been generated to describe each word in the feature space by extracting unıform zoning-, diagonal- and centroid-based features from the database of handwritten word images (50-word classes). Various classifiers, namely K-nearest neighbor (KNN), decision tree and random forest, are employed for the recognition purpose. Furthermore to enhance the system performance, combination of above mentioned features along with gradient boosted decision tree algorithm is proposed. In this way, proposed system achieved maximum recognition accuracy of 94.53% and the attained competent results are comparable with exiting state-of-the-art methods. Moreover, the proposed system has achieved F1-score of 94.56%, FAR of 0.11%, FRR of 5.46%, MCC of 0.945 and AUC of 97.21%. [ABSTRACT FROM AUTHOR]

This paper presents an efficient skull stripping method to improve the decision-making process. Extended Weiner filtering (EWF) is used for removing the noise and enhancing the quality of images. Further, laplacian lion optimization algorithm (LXLOA) is implemented. LXLOA utilizes the Otsu's and Tsallis entropy fitness function to determine an optimal solution. The implemented LXLOA provides a threshold value required for performing the segmentation on the brain MRI images. The extracted features are selected using fuzzy weighted k-means embedding LDA (linear discriminant analysis) method for improving training of the classification model. The proposed LXLOA is extensively tested on standard benchmark functions CEC 2017 and outperforms the existing state-of-the-art algorithm. Rigorous statistical analysis is conducted to determine the statistical significance. Three-fold performance comparison is performed by considering (a) the quality of the segmented image; (b) accuracy, sensitivity, and specificity; and (c) computational cost of convergence for finding an optimal solution. Result reveals that LXLOA gives promising results and demonstrate effective outcomes on the standard quality measures (a) accuracy (97.37%); (b) sensitivity (85.8%); (c) specificity (90%); and (d) precision (91.92%). [ABSTRACT FROM AUTHOR]

Breast cancer, which is also the leading cause of death among women, is one of the most common forms of the disease that affects females all over the world. The discovery of breast cancer at an early stage is extremely important because it allows selecting appropriate treatment protocol and thus, stops the development of cancer cells. In this paper, a new patients detection strategy has been presented to identify patients with the disease earlier. The proposed strategy composes of two parts which are data preprocessing phase and patient detection phase (PDP). The purpose of this study is to introduce a feature selection methodology for determining the most efficient and significant features for identifying breast cancer patients. This method is known as new hybrid feature selection method (NHFSM). NHFSM is made up of two modules which are quick selection module that uses information gain, and feature selection module that uses hybrid bat algorithm and particle swarm optimization. Consequently, NHFSM is a hybrid method that combines the advantages of bat algorithm and particle swarm optimization based on filter method to eliminate many drawbacks such as being stuck in a local optimal solution and having unbalanced exploitation. The preprocessed data are then used during PDP in order to enable a quick and accurate detection of patients. Based on experimental results, the proposed NHFSM improves the efficiency of patients' classification in comparison with state-of-the-art feature selection approaches by roughly 0.97, 0.76, 0.75, and 0.716 in terms of accuracy, precision, sensitivity/recall, and F-measure. In contrast, it has the lowest error rate value of 0.03. [ABSTRACT FROM AUTHOR]

Spiking neural networks (SNNs) are mimicking computationally powerful biologically inspired models in which neurons communicate through sequences of spikes, regarded here as sparse binary sequences of zeros and ones. In neuroscience it is conjectured that time encoding, where the information is carried by the temporal position of spikes, is playing a crucial role at least in some parts of the brain where estimation of the spiking rate with a large latency cannot take place. Motivated by the efficiency of temporal coding, compared with the widely used rate coding, the goal of this paper is to develop and train an energy-efficient time-coded deep spiking neural network system. To ensure that the similarity among input stimuli is translated into a correlation of the spike sequences, we introduce correlative temporal encoding and extended correlative temporal encoding techniques to map analog input information into input spike patterns. Importantly, we propose an implementation where all multiplications in the system are replaced with at most a few additions. As a more efficient alternative to both rate-coded SNNs and artificial neural networks, such system represents a preferable solution for the implementation of neuromorphic hardware. We consider data classification tasks where input spike patterns are presented to a feed-forward architecture with leaky-integrate-and-fire neurons. The SNN is trained by backpropagation through time with the objective to match sequences of output spikes with those of specifically designed target spike patterns, each corresponding to exactly one class. During inference the target spike pattern with the smallest van Rossum distance from the output spike pattern determines the class. Extensive simulations indicate that the proposed system achieves a classification accuracy at par with that of state-of-the-art machine learning models. [ABSTRACT FROM AUTHOR]