Search

Nowadays, automatic deception detection has received considerable attention in the machine learning community owing to this research interest to its vast applications in the fields of social media, interviews, law enforcement, and the military. In this study, a novel deep convolution neural network (DCNN) named LieNet is proposed to precisely detect the multi-scale variations of deception automatically. Our approach is a combination of contact and non-contact-based approaches. First, 20 frames from each video are fetched and concatenated to form a single image. Moreover, an audio signal is extracted from video and treated as image input by plotting the signal into two dimensional (2D) plane. Furthermore, 13 channels of EEG signals are plotted into 2D plane and concatenated to generate an image. Second, the LieNet model extracts features from each modality separately. Third, scores are estimated using a softmax classifier for all the modalities. Finally, three scores are combined using score level fusion to obtain a score, which gives support in favor of either deception or truth. The LieNet is validated on the “Bag-of-Lies (BoL)”“, Real-life (RL) trail”, and “Miami University Deception Detection (MU3D)" databases by considering four evaluation indexes viz., accuracy, precision, recall, and F1-score. Experimental outcomes depict that the LieNet defeats an initial work on Set-A and Set-B of the BoL database with average accuracies of 95.91% and 96.04% respectively. The accuracies obtained by the LieNet are 97% and 98% on RL trail and MU3D databases respectively.

This paper proposes a novel local feature descriptor coined as a local instant-and-center-symmetric neighbor-based pattern of the extrema-images (LINPE) to detect breast abnormalities in thermal breast images. It is a hybrid descriptor that combines two different feature descriptors: one is the inverse-probability difference extrema (IpDE), and another is the local instant and center-symmetric neighbor-based pattern (LICsNP). IpDE is developed to compute the intensity-inhomogeneity-invariant feature-based image of the breast thermogram. Besides, the LICsNP is intended to capture the local microstructure pattern information in the IpDE image. A new paradigm, named Broad Learning (BL) network, is introduced here as a classifier to differentiate the healthy and sick breast thermograms efficiently. The efficacy of the proposed system is quantitatively validated on the images of DMR-IR and DBT-TU-JU databases. Extensive experimentation on these databases with an average accuracy of 96.90% and 94%, respectively, justifies proposed system's superiority in the differentiation of healthy and sick breast thermograms over the other related existing state-of-the-art methods. The proposed system also performs consistently in the presence of noise and rotational changes.

To understand the connection among the brain and facial muscles, we decoded the correlation of facial muscles and brain activities. We applied information theory (Shannon entropy) to EEG and EMG si...

Automatic facial expression recognition (FER) is one of the most challenging tasks in computer vision. FER admits a wide range of applications in human–computer interaction, behavioral psychology, and human expression synthesis. Extensive works have been reported in this field, mainly, based on handcrafted features. However, it is a challenging task to accurately extract all the correlated handcrafted features due to the effect of variations caused by emotional state. Therefore, there is a quest for further research on accurately extracting relevant features that can capture changes in facial expressions (FEs) with high fidelity. In this study, we propose FER-net: a convolution neural network to distinguish FEs efficiently with the help of the softmax classifier. We implement our method FER-net along with twenty-one state-of-the-art methods and test them on five benchmarking datasets, namely FER2013, Japanese Female Facial Expressions, Extended CohnKanade, Karolinska Directed Emotional Faces, and Real-world Affective Faces. Seven FEs, namely neutral, anger, disgust, fear, happiness, sadness, and surprise, are considered in this work. The average accuracies on these datasets are 78.9%, 96.7%, 97.8%, 82.5% and 81.68%, respectively. The obtained results demonstrate that FER-net is preeminent in comparison with twenty-one state-of-the-art methods.

In 2019, citizens of Indonesia participated in the democratic process of electing a new president, vice president, and various legislative candidates for the country. The 2019 Indonesian presidential election was very tense in terms of the candidates' campaigns in cyberspace, especially on social media sites such as Facebook, Twitter, Instagram, Google+, Tumblr, LinkedIn, etc. The Indonesian people used social media platforms to express their positive, neutral, and also negative opinions on the respective presidential candidates. The campaigning of respective social media users on their choice of candidates for regents, governors, and legislative positions up to presidential candidates was conducted via the Internet and online media. Therefore, the aim of this paper is to conduct sentiment analysis on the candidates in the 2019 Indonesia presidential election based on Twitter datasets. The study used datasets on the opinions expressed by the Indonesian people available on Twitter with the hashtags (#) containing "Jokowi and Prabowo." We conducted data pre-processing using a selection of comments, data cleansing, text parsing, sentence normalization and tokenization based on the given text in the Indonesian language, determination of class attributes, and, finally, we classified the Twitter posts with the hashtags (#) using Naïve Bayes Classifier (NBC) and a Support Vector Machine (SVM) to achieve an optimal and maximum optimization accuracy. The study provides benefits in terms of helping the community to research opinions on Twitter that contain positive, neutral, or negative sentiments. Sentiment Analysis on the candidates in the 2019 Indonesian presidential election on Twitter using non-conventional processes resulted in cost, time, and effort savings. This research proved that the combination of the SVM machine learning algorithm and alphabetic tokenization produced the highest accuracy value of 79.02%. While the lowest accuracy value in this study was obtained with a combination of the NBC machine learning algorithm and N-gram tokenization with an accuracy value of 44.94%. ABSTRAK: Pada tahun 2019 rakyat Indonesia telah terlibat dalam proses demokrasi memilih presiden baru, wakil presiden, dan berbagai calon legislatif negara. Pemilihan presiden Indonesia 2019 sangat tegang dalam kempen calon di ruang siber, terutama di laman media sosial seperti Facebook, Twitter, Instagram, Google+, Tumblr, LinkedIn, dll. Rakyat Indonesia menggunakan platfom media sosial bagi menyatakan pendapat positif, berkecuali, dan juga negatif terhadap calon presiden masing-masing. Kampen pencalonan menteri, gabenor, dan perundangan hingga pencalonan presiden dilakukan melalui media internet dan atas talian. Oleh itu, kajian ini dilakukan bagi menilai sentimen terhadap calon pemilihan presiden Indonesia 2019 berdasarkan kumpulan data Twitter. Kajian ini menggunakan kumpulan data yang diungkapkan oleh rakyat Indonesia yang terdapat di Twitter dengan hashtag (#) yang mengandungi "Jokowi dan Prabowo." Proses data dibuat menggunakan pilihan komentar, pembersihan data, penguraian teks, normalisasi kalimat, dan tokenisasi teks dalam bahasa Indonesia, penentuan atribut kelas, dan akhirnya, pengklasifikasian catatan Twitter dengan hashtag (#) menggunakan Klasifikasi Naïve Bayes (NBC) dan Mesin Vektor Sokongan (SVM) bagi mencapai ketepatan optimum dan maksimum. Kajian ini memberikan faedah dari segi membantu masyarakat meneliti pendapat di Twitter yang mengandungi sentimen positif, neutral, atau negatif. Analisis Sentimen terhadap calon dalam pemilihan presiden Indonesia 2019 di Twitter menggunakan proses bukan konvensional menghasilkan penjimatan kos, waktu, dan usaha. Penyelidikan ini membuktikan bahawa gabungan algoritma pembelajaran mesin SVM dan tokenisasi abjad menghasilkan nilai ketepatan tertinggi iaitu 79.02%. Manakala nilai ketepatan terendah dalam kajian ini diperoleh dengan kombinasi algoritma pembelajaran mesin NBC dan tokenisasi N-gram dengan nilai ketepatan 44.94%.

Depression is a common reason for an increase in suicide cases worldwide. Thus, to mitigate the effects of depression, accurate diagnosis and treatment are needed. An electroencephalogram (EEG) is an instrument used to measure and record the brain’s electrical activities. It can be utilized to produce the exact report on the level of depression. Previous studies proved the feasibility of the usage of EEG data and deep learning (DL) models for diagnosing mental illness. Therefore, this study proposes a DL-based convolutional neural network (CNN) called DeprNet for classifying the EEG data of depressed and normal subjects. Here, the Patient Health Questionnaire 9 score is used for quantifying the level of depression. The performance of DeprNet in two experiments, namely, the recordwise split and the subjectwise split, is presented in this study. The results attained by DeprNet have an accuracy of 0.9937, and the area under the receiver operating characteristic curve (AUC) of 0.999 is achieved when recordwise split data are considered. On the other hand, an accuracy of 0.914 and the AUC of 0.956 are obtained, while subjectwise split data are employed. These results suggest that CNN trained on recordwise split data gets overtrained on EEG data with a small number of subjects. The performance of DeprNet is remarkable compared with the other eight baseline models. Furthermore, on visualizing the last CNN layer, it is found that the values of right electrodes are prominent for depressed subjects, whereas, for normal subjects, the values of left electrodes are prominent.

Nowadays, the usage of deep learning-based approaches for salient object detection (SOD) is increasing exponentially to detect and localize visually distinct regions in static images. However, the variability in scales of salient objects requires further attention given the abstract nature of the multilayer feature hierarchy of convolution neural networks (CNNs). First, feature maps of different layers in CNNs embed abstract information about objects that changes with the object’s scale. Second, the progressive feature fusion in models, such as the feature pyramid network, loses its effectiveness in detecting sharp boundaries due to the late fusion of detailed features. This work proposes two modules namely, adjacent layer attention block and partial encoder–decoder (PED) block to handle the aforementioned issues. The proposed adjacent layer attention block facilitates communication among the layers of closest abstraction to mine abundant scale features at the current resolution. The resultant integrated feature at a resolution contains detailed and semantic information from interaction among the adjacent layers useful to extract scale information of complex objects. A PED module utilizes the resolution-specific integrated features from the adjacent layer attention block of its corresponding encoder to generate multiscale features, and fuse them in a top-down manner. This level-specific distribution of aggregated features within a PED helps coarser layers of the network to acquire boundary information. Experimental results on five broadly used SOD datasets are compared with recent 20 state-of-the-art SOD models. The proposed method performs favorably against its competitors without any preprocessing or postprocessing.

The authors wish to thank Universiti Teknologi Malaysia (UTM) for its support under Research University Grant Vot- 20H04, Malaysia Research University Network (MRUN) Vot 4L876. The authors would like to acknowledge that this work was supported/funded by the Ministry of Higher Education under the Fundamental Research Grant Scheme (FRGS/1/2018/ICT04/UTM/01/1). The work was also partially supported by the Specific Research project (SPEV) at the Faculty of Informatics and Management, University of Hradec Kralove, Czech Republic, under Grant 2102-2021. The authors are grateful for the support of student Sebastien Mambou in consultations regarding application aspects. The authors also wish to thank the Ministry of Education Malaysia for the Hadiah Latihan Persekutuan (HLP) scholarship to complete the research., A botnet is a malware program that a hacker remotely controls called a botmaster. Botnet can perform massive cyber-attacks such as DDOS, SPAM, click-fraud, information, and identity stealing. The botnet also can avoid being detected by a security system. The traditional method of detecting botnets commonly used signature-based analysis unable to detect unseen botnets. The behavior-based analysis seems like a promising solution to the current trends of botnets that keep evolving. This paper proposes a multilayer framework for botnet detection using machine learning algorithms that consist of a ltering module and classi cation module to detect the botnet's command and control server. We highlighted several criteria for our framework, such as it must be structure-independent, protocol-independent, and able to detect botnet in encapsulated technique. We used behavior-based analysis through ow-based features that analyzed the packet header by aggregating it to a 1-s time. This type of analysis enables detection if the packet is encapsulated, such as using a VPN tunnel. We also extend the experiment using different time intervals, but a 1-s time interval shows the most impressive results. The result shows that our botnet detection method can detect up to 92% of the f-score, and the lowest false-negative rate was 1.5%., Universiti Teknologi Malaysia (UTM) through the Research University Vot-20H04, Malaysia Research University Network (MRUN) Vot4L876, Ministry of Higher Education through the Fundamental Research Grant Scheme FRGS/1/2018/ICT04/UTM/01/1, Hadiah Latihan Persekutuan (HLP) Scholarship through the Ministry of Education Malaysia, Specific Research Project (SPEV) by the Faculty of Informatics and Management, University of Hradec Kralove, Czech Republic

Computer-aided diagnosis of disease primarily depends on proper vision-based measurement (VBM). The traditional approach followed for diagnosis of colorectal cancer includes a manual screening of colorectum via a colonoscope and resection of polyps for histopathological analysis to decide the grade of malignancy. This procedure is time-consuming and expensive, and removal of benign polyp for analysis signifies the inefficiency of the diagnosis system. These drawbacks inspired us to develop an automatic vision-based analysis method for preliminary in vivo malignancy analysis of the polyp region. In this work, we have proposed a fusion-based polyp segmentation network, namely, Polyp-Net. Recently, convolutional neural networks (CNNs) have shown immense success in the domain of medical image analysis as it can exploit in-depth significant features with high discrimination capabilities. Therefore, motivated by these insights, we have proposed an enriched version of CNN with a nascent pooling mechanism, namely dual-tree wavelet pooled CNN (DT-WpCNN). The resultant segmented mask contains some surplus high-intensity regions apart from the polyp region. These shortcomings are avoided using a new variation of the region-based level-set method, namely, the local gradient weighting-embedded level-set method (LGWe-LSM), which shows a significant reduction of false-positive rate. The pixel-level fusion of the two enhanced methods shows more potentiality in the segmentation of the polyp regions. Our proposed network is trained on CVC-colon DB and tested on CVC-clinic DB. It achieves a dice score of 0.839, volume-similarity of 0.863, precision of 0.836, recall of 0.811, F1-score of 0.823, F2-score of 0.815, and Hausdorff distance of 21.796 which outperforms the existing baseline CNN’s and recent state-of-the-art methods.

The effective extraction of local and contextual visual cues carrying information of different scales is crucial for accurate detection of the salient object(s) with varying shape, size, and location. The atrous spatial pyramid pooling (ASPP) and its dense versions are widely used for extracting contextual features for dense prediction tasks. The skip connections in densely or moderately connected ASPP directly propagate the context information from a parallel dilated convolution to the next higher rate dilated convolution to combat the “gridding issue” in atrous convolutions. The aggregated context from several scales may dilute features belonging to small objects or confuse between the salient object and the background. To emphasize invariance features for different scale visual patterns in an image, a gate-based context extraction module is proposed in this work. Gate functions are embedded in the interbranch short connection of the proposed module. The learnable gates are deployed to decide on the relevance of the contextual information extracted at a lower scale for the next higher scale. Experimental results on salient object detection tasks demonstrate that gates are helpful to retain relevant contextual information across multiple-scales of the context-extraction module. The performance of the proposed gated contextual feature-based salient object detector is evaluated on five broadly used saliency detection benchmarks by comparing it with the other 13 state-of-the-art approaches. Experimental outcomes show that the proposed method achieves a favorable performance for various compared evaluation measures.

Clustering inspired superpixel algorithms perform a restricted partitioning of an image, where each visually coherent region containing perceptually similar pixels serves as a primitive in subsequent processing stages. Simple linear iterative clustering (SLIC) has emerged as a standard superpixel generation tool due to its exceptional performance in terms of segmentation accuracy and speed. However, SLIC applies a manually adjusted distance measure for dis-similarity computation which directly affects the quality of superpixels. In this work, self-adjustable distance measures are adapted from the weighted k-means clustering (W-k-means) for generating superpixel segmentation. In the proposed distance measures, an adaptive weight associated with each variable reflects its relevance in the clustering process. Intuitively, the variable weights correspond to the normalization terms in SLIC that affect the trade-off between superpixels boundary adherence and compactness. Weights that influence consistency in superpixel generation are automatically updated. The variable weights update is accomplished during optimization with a closed-form solution based on the current image partition. The proposed adaptive, W-k-means-based superpixels (AWkS) experimented on three benchmarks under different distance measure outperform the conventional SLIC algorithm with respect to various boundary adherence metrics. Finally, the effectiveness of the AWkS over SLIC is demonstrated for saliency detection.

This paper proposes a mathematical model for decomposing a range face image into four basic components (named ‘complement components’) in conjunction with a simple approach for data-level fusion to generate thirty-six additional hybrid components. These forty component faces composing a new face image space called the ‘complement component face space.’ The main challenge of this work was to extract relevant features from the vast face space. Features are extracted from the four basic components and four selected hybrid components using singular value decomposition. To introduce diversity, the extracted feature vectors are fused by applying the crossover operation of the genetic algorithm using a Hamming distance-based fitness measure. Particle swarm optimization-based feature selection is employed on the fused features to discard redundant feature values and to maximize the face recognition performance. The recognition performances of the proposed feature set with a support vector machine-based classifier on three accessible and well-known 3D face databases, namely, Frav3D, Bosphorus, and Texas3D, show significant improvements over those achieved by state-of-the-art methods. This work also studies the feasibility of utilizing the component images in the complement component face space for data augmentation in convolutional neural network (CNN)-based frameworks.

The measurement of the surface quality and the profile preciseness is major issues in many industrial branches such that the surface quality of semi products directly affects the subsequent production steps. Although, there are many ways to obtain required data, the hardware necessary for the measurements such as 2D or 3D scanners, depending on the problem’s complexity, is too expensive. Therefore, in this paper, what we put forward as a novelty is an algorithm which is verified on the model of simple 3D scanner on the image processing basis with the resolution of 0.1 mm. There are many ways to scan surface profile; however, the image processing currently is the most trending topic in industry automation. Most importantly, in order to obtain surface images, standard high resolution reflex camera is used and thus the post processing could be realized with MatLab as the software environment. Therefore, this solution is an alternative to the expensive scanners, and single-purpose devices could be extended by many additional functions.

In recent years, the number of image fusion schemes presented by the research community has increased significantly. Measuring the performance of these schemes is an important issue. In this work, we introduce three quantitative fusion metrics to assess the quality of an image fusion algorithm. The proposed metrics rely on edge information that is obtained using fractional order differentiation. Edge and orientation strengths are fed into three sigmoidal functions separately for estimating the values of three normalized weighted metrics for the fused image corresponding to source images. The experiments on the multi-focus, infrared-visible and medical image fusion pairs demonstrate that the proposed fusion metrics are perceptually meaningful and outperform some of the state-of-the-art metrics. This work was supported in part by the project (Prediction of diseases through computer assisted diagnosis system using images captured by minimally-invasive and non-invasive modalities), Computer Science and Engineering, PDPM Indian Institute of Information Technology, Design and Manufacturing, Jabalpur, India, under Grant SPARC-MHRD-231, in part by the project of Grant Agency of Excellence, University of Hradec Kralove, Faculty of Informatics and Management, Czech Republic, under Grant UHK-FIMGE-2020, and in part by the IT4Neuro—project of the Ministry of Education, Youth and Sports of Czech Republic under Project ERDF CZ.02.1.01/0.0/0.0/18 _069/0010054.

This work was supported/funded by the Ministry of Higher Education under the Fundamental Research Grant Scheme (FRGS/1/2018/ICT04/UTM/01/1). The authors sincerely thank Universiti Teknologi Malaysia (UTM) under Research University Grant Vot-20H04, Malaysia Research University Network (MRUN) Vot 4L876, for the completion of the research. Faculty of Informatics and Management, University of Hradec Kralove, SPEV project Grant Number: 2102/2021., Phishing detection with high-performance accuracy and low computational complexity has always been a topic of great interest. New technologies have been developed to improve the phishing detection rate and reduce computational constraints in recent years. However, one solution is insufficient to address all problems caused by attackers in cyberspace. Therefore, the primary objective of this paper is to analyze the performance of various deep learning algorithms in detecting phishing activities. This analysis will help organizations or individuals select and adopt the proper solution according to their technological needs and specific applications’ requirements to fight against phishing attacks. In this regard, an empirical study was conducted using four different deep learning algorithms, including deep neural network (DNN), convolutional neural network (CNN), Long Short-Term Memory (LSTM), and gated recurrent unit (GRU). To analyze the behaviors of these deep learning architectures, extensive experiments were carried out to examine the impact of parameter tuning on the performance accuracy of the deep learning models. In addition, various performance metrics were measured to evaluate the effectiveness and feasibility of DL models in detecting phishing activities. The results obtained from the experiments showed that no single DL algorithm achieved the best measures across all performance metrics. The empirical findings from this paper also manifest several issues and suggest future research directions related to deep learning in the phishing detection domain., Ministry of Higher Education under the Fundamental Research Grant Scheme FRGS/1/2018/ICT04/UTM/01/1, Universiti Teknologi Malaysia (UTM) Vot-20H04, Malaysia Research University Network (MRUN) 4L876, Faculty of Informatics and Management, University of Hradec Kralove, SPEV project 2102/2021.

A considerable amount of research has been done since long to select an appropriate similarity or dissimilarity measure in cluster analysis for exposing the natural grouping in an input dataset. Still, it is an open problem. In recent years, the research community is focusing on divergence-based non-Euclidean similarity measure in partitional clustering for grouping. In this paper, the Euclidean distance of traditional Fuzzy k-means (FKM) algorithm is replaced by the S-distance, which is derived from the newly introduced S-divergence. Few imperative properties of S-distance and modified FKM are presented in this study. The performance of the proposed FKM is compared with the conventional FKM with Euclidean distance and its variants with the help of several synthetic and real-world datasets. This study focuses on how the proposed clustering algorithm performs on the adopted datasets empirically. The comparative study illustrates that the obtained results are convincing. Moreover, the achieved results denote that the modified FKM outperforms some state-of-the-art FKM algorithms.

Missing rainfall data have reduced the quality of hydrological data analysis because they are the essential input for hydrological modeling. Much research has focused on rainfall data imputation. However, the compatibility of precipitation (rainfall) and non-precipitation (meteorology) as input data has received less attention. First, we propose a novel input structure for the missing data imputation method. Principal component analysis (PCA) is used to extract the most relevant features from the meteorological data. This paper introduces the combined input of the significant principal components (PCs) and rainfall data from nearest neighbor gauging stations as the input to the estimation of the missing values. Second, the effects of the combination input for infilling the missing rainfall data series were compared using the sine cosine algorithm neural network (SCANN) and feedforward neural network (FFNN). The results showed that SCANN outperformed FFNN imputation in terms of mean absolute error (MAE), root means square error (RMSE) and correlation coefficient (R), with an average accuracy of more than 90%. This study revealed that as the percentage of missingness increased, the precision of both imputation methods reduced.

The use of long short-term memory (LSTM) for session-based recommendations is described in this research. This study uses char-level LSTM as a real-time recommendation service to test and offer the optimal solution. Our strategy can be used to any situation. Two LSTM layers and a thick layer make up our model. To evaluate the prediction results, we use the mean of squared errors. We also put our recall and precision metrics prediction to the test. The best-performing network had roughly 2000 classes and was a trainer for the last year of likes on an image-based social platform. On twenty objects, our best model had a recall value of 0.182 and a precision value of 0.061.

The rapid growth of emerging technologies, smart devices, 5G communication, etc. have contributed to the accumulation of data, hence introducing the big data era. Big data imposes a variety of challenges associated with machine learning, especially in phishing detection. Therefore, this paper aims to provide an analysis and summary of current research in phishing detection through machine learning for big data. To achieve this goal, this study adopted a systematic literature review (SLR) technique and critically analyzed a total of 30 papers from various journals and conference proceedings. These papers were selected from previous studies in five different databases on content published between 2018 and January 2021. The results obtained from this study reveal a limited number of research works that comprehensively reviewed the feasibility of applying both machine learning and big data technologies in the context of phishing detection.

Each year, malware issues remain one of the cybersecurity concerns since malware’s complexity is constantly changing as the innovation rapidly grows. As a result, malware attacks have affected everyday life from various mediums and ways. Therefore, a machine learning algorithm is one of the essential solutions in the security of computer systems to detect malware regarding the ability of machine learning algorithms to keep up with the evolution of malware. This paper is devoted to reviewing the most up-to-date research works from 2017 to 2021 on malware detection where machine learning algorithm including K-Means, Decision Tree, Meta-Heuristic, Naive Bayes, Neuro-fuzzy, Bayesian, Gaussian, Support Vector Machine (SVM), K-Nearest Neighbour (KNN) and n-Grams was discovered using a systematic literature review. This paper aims at the following: (1) it describes each machine learning algorithm, (2) for each algorithm; it shows the performance of malware detection, and (3) we present the challenges and limitations of the algorithm during research processes.

The increasing popularity of social networks indicates that the vast amounts of data contained within them could be useful in various implementations, including recommendation systems. Interests and research publications on deep learning-based recommendation systems have largely increased. This study aimed to identify, summarize, and assess studies related to the application of deep learning-based recommendation systems on social media platforms to provide a systematic review of recent studies and provide a way for further research to improve the development of deep learning-based recommendation systems in social environments. A total of 32 papers were selected from previous studies in five of the major digital libraries, including Springer, IEEE, ScienceDirect, ACM, Scopus, and Web of Science, published between 2016 and 2020. Results revealed that even though RS has received high coverage in recent years, several obstacles and opportunities will shape the future of RS for researchers. In addition, social recommendation systems achieving high accuracy can be built by using a combination of techniques that incorporate a range of features in SRS. Therefore, the adoption of deep learning techniques in developing social recommendation systems is undiscovered.

Education is one of the important fields in this challenging world. The researchers come out with the new perceptive, which is learning analytics that is a new invention for helping out the instructors, learners, and administrators. The use of learning analytics can be the medium for increasing the productivity of education for producing capable leaders in the future. Machine learning comes out with any type of techniques such as Decision Tree, Support Vector Machine, Naive Bayes, and Ensemble Classifiers. However, both Decision Tree and Ensemble Classifiers are chosen as the best potential machine learning techniques to cope with the large database of students. The Boosted Tree of Ensemble Classifiers managed to get 99.6% accuracy of training 378,005 data of students regarding the Virtual Learning Environment (VLE).

This study proposes an alternate data extraction method that combines three well-known feature selection methods for handling large and problematic datasets: the correlation-based feature selection (CFS), best first search (BFS), and dominance-based rough set approach (DRSA) methods. This study aims to enhance the classifier’s performance in decision analysis by eliminating uncorrelated and inconsistent data values. The proposed method, named CFS-DRSA, comprises several phases executed in sequence, with the main phases incorporating two crucial feature extraction tasks. Data reduction is first, which implements a CFS method with a BFS algorithm. Secondly, a data selection process applies a DRSA to generate the optimized dataset. Therefore, this study aims to solve the computational time complexity and increase the classification accuracy. Several datasets with various characteristics and volumes were used in the experimental process to evaluate the proposed method’s credibility. The method’s performance was validated using standard evaluation measures and benchmarked with other established methods such as deep learning (DL). Overall, the proposed work proved that it could assist the classifier in returning a significant result, with an accuracy rate of 82.1% for the neural network (NN) classifier, compared to the support vector machine (SVM), which returned 66.5% and 49.96% for DL. The one-way analysis of variance (ANOVA) statistical result indicates that the proposed method is an alternative extraction tool for those with difficulties acquiring expensive big data analysis tools and those who are new to the data analysis field., Ministry of Higher Education under the Fundamental Research Grant Scheme (FRGS/1/2018/ICT04/UTM/01/1), Universiti Teknologi Malaysia (UTM) under Research University Grant Vot-20H04, Malaysia Research University Network (MRUN) Vot 4L876, SPEV project, University of Hradec Kralove, Faculty of Informatics and Management, Czech Republic (ID: 2102–2021), “Smart Solutions in Ubiquitous Computing Environments”

This work was supported in part by the Ministry of Higher Education through the Fundamental Research Scheme under Grant FRGS/1/2018/ICT04/UTM/01/1, in part by the Speci~c Research Project (SPEV) at the Faculty of Informatics and Management, University of Hradec Kralove, Czech Republic, under Grant 2102-2021, in part by the Universiti Teknologi Malaysia (UTM) under Research University Grant Vot-20H04, and in part by the Malaysia Research University Network (MRUN) under Grant Vot 4L876., Today, predictive analytics applications became an urgent desire in higher educational institutions. Predictive analytics used advanced analytics that encompasses machine learning implementation to derive high-quality performance and meaningful information for all education levels. Mostly know that student grade is one of the key performance indicators that can help educators monitor their academic performance. During the past decade, researchers have proposed many variants of machine learning techniques in education domains. However, there are severe challenges in handling imbalanced datasets for enhancing the performance of predicting student grades. Therefore, this paper presents a comprehensive analysis of machine learning techniques to predict the nal student grades in the rst semester courses by improving the performance of predictive accuracy. Two modules will be highlighted in this paper. First, we compare the accuracy performance of six well-known machine learning techniques namely Decision Tree (J48), Support Vector Machine (SVM), Naïve Bayes (NB), K-Nearest Neighbor (kNN), Logistic Regression (LR) and Random Forest (RF) using 1282 real student's course grade dataset. Second, we proposed a multiclass prediction model to reduce the over tting and misclassi cation results caused by imbalanced multi-classi cation based on oversampling Synthetic Minority Oversampling Technique (SMOTE) with two features selection methods. The obtained results showthat the proposed model integrates with RF give signi cant improvement with the highest f-measure of 99.5%. This proposed model indicates the comparable and promising results that can enhance the prediction performance model for imbalanced multi-classi cation for student grade prediction., Science and Technology Development Fund (STDF), Ministry of Higher Education & Scientific Research (MHESR) FRGS/1/2018/ICT04/UTM/01/1, Specific Research Project (SPEV) at the Faculty of Informatics and Management, University of Hradec Kralove, Czech Republic 2102-2021, Universiti Teknologi Malaysia (UTM) Vot-20H04, Malaysia Research University Network (MRUN) 4L876

Student performance is the most factor that can be beneficial for many parties, including students, parents, instructors, and administrators. Early prediction is needed to give the early monitor by the responsible person in charge of developing a better person for the nation. In this paper, the improvement of Bagged Tree to predict student performance based on four main classes, which are distinction, pass, fail, and withdrawn. The accuracy is used as an evaluation parameter for this prediction technique. The Bagged Tree with the addition of Bag, AdaBoost, RUSBoost learners helps to predict the student performance with the massive datasets. The use of the RUSBoost algorithm proved that it is very suitable for the imbalance datasets as the accuracy is 98.6% after implementing the feature selection and 99.1% without feature selection compared to other learner types even though the data is more than 30,000 datasets.

Normative multi-agent research is an alternative viewpoint in the design of adaptive autonomous agent architecture. Norms specify the standards of behaviors such as which actions or states should be achieved or avoided. The concept of norm synthesis is the process of generating useful normative rules. This study proposes a model for normative rule extraction from implicit learning, namely using the Q-learning algorithm, into explicit norm representation by implementing Dynamic Deontics and Hierarchical Knowledge Base (HKB) to synthesize useful normative rules in the form of weighted state-action pairs with deontic modality. OpenAi Gym is used to simulate the agent environment. Our proposed model is able to generate both obligative and prohibitive norms as well as deliberate and execute said norms. Results show the generated norms are best used as prior knowledge to guide agent behavior and performs poorly if not complemented by another agent coordination mechanism. Performance increases when using both obligation and prohibition norms, and in general, norms do speed up optimum policy reachability.

Data-driven plays an important role in determining the quality of services in institutions of higher learning (HEIs). Increasingly data in education is encouraging institutions to find ways to improve student academic performance. By using machine learning with visual analytics, data can be predicted based on valuable information and presented with interactive visualizations for institutions to improve decision making. Therefore, predicting students’ academic performance is critical to identifying students at risk of failing a course. In this paper, we propose two approaches, such as (i) a prediction model for predicting students’ final grade based on machine learning that interacts with computational models; (ii) visual analytics to visualize predictive models and insightful data for educators. The data were tested using student achievement records collected from one of the Malaysian Polytechnic databases. The data set used in this study involved 489 first semester students in Computer System Architecture (CSA) course from 2016 to 2019. The decision tree algorithms (J48), Random Tree (RT), Random Forest (RF), and REPTree) was used on the student data set to produce the best predictions of the model. Experimental results show that J48 returns the highest accuracy with 99.8 %, among other algorithms. The findings of this study can help educators predict student success or failure for a particular course at the end of the semester and help educators make informed decisions to improve student academic performance at Polytechnic Malaysia.

One of the most recent non-invasive technologies to examine the gastrointestinal tract is wireless capsule endoscopy (WCE). As there are thousands of endoscopic images in an 8–15 h long video, an evaluator has to pay constant attention for a relatively long time (60–120 min). Therefore the possibility of the presence of pathological findings in a few images (displayed for evaluation for a few seconds only) brings a significant risk of missing the pathology with all negative consequences for the patient. Hence, manually reviewing a video to identify abnormal images is not only a tedious and time consuming task that overwhelms human attention but also is error prone. In this paper, a method is proposed for the automatic detection of abnormal WCE images. The differential box counting method is used for the extraction of fractal dimension (FD) of WCE images and the random forest based ensemble classifier is used for the identification of abnormal frames. The FD is a well-known technique for extraction of features related to texture, smoothness, and roughness. In this paper, FDs are extracted from pixel-blocks of WCE images and are fed to the classifier for identification of images with abnormalities. To determine a suitable pixel block size for FD feature extraction, various sizes of blocks are considered and are fed into six frequently used classifiers separately, and the block size of 7 × 7 giving the best performance is empirically determined. Further, the selection of the random forest ensemble classifier is also done using the same empirical study. Performance of the proposed method is evaluated on two datasets containing WCE frames. Results demonstrate that the proposed method outperforms some of the state-of-the-art methods with AUC of 85 % and 99 % on Dataset-I and Dataset-II respectively.

Multi-focus image fusion is defined as “the combination of a group of partially focused images of a same scene with the objective of producing a fully focused image.” Normally, transform-domain-based image fusion methods preserve the textures and edges in the blend image, but many are translation variant. The translation-invariant transforms produce the same size approximation and detail images, which are more convenient to devise the fusion rules. In this work, a translation-invariant multi-focus image fusion approach using the à-trous wavelet transform is introduced, which uses fractal dimension as a clarity measure for the approximation coefficients and Otsu’s threshold to fuse the detail coefficients. The subjective assessment of the proposed method is carried out using the fusion results of nine state-of-the-art methods. On the other hand, eight fusion quality metrics are considered for the objective assessment. The results of subjective and objective assessment on grayscale and color multi-focus image pairs illustrate that the proposed method is competitive and even better than some of the existing methods.

Keywords: crypto-ransomware, locker-ransomware, static analysis, dynamic analysis, machine learning.

Background: In biomedicine, infrared thermography is the most promising technique among other conventional methods for revealing the differences in skin temperature, resulting from the irregular temperature dispersion, which is the significant signaling of diseases and disorders in human body. Given the process of detecting emitted thermal radiation of human body temperature by infrared imaging, we, in this study, present the current utility of thermal camera models namely FLIR and SEEK in biomedical applications as an extension of our previous article. Results: The most significant result is the differences between image qualities of the thermograms captured by thermal camera models. In other words, the image quality of the thermal images in FLIR One is higher than SEEK Compact PRO. However, the thermal images of FLIR One are noisier than SEEK Compact PRO since the thermal resolution of FLIR One is 160 × 120 while it is 320 × 240 in SEEK Compact PRO. Conclusion: Detecting and revealing the inhomogeneous temperature distribution on the injured toe of the subject, we, in this paper, analyzed the imaging results of two different smartphone-based thermal camera models by making comparison among various thermograms. Utilizing the feasibility of the proposed method for faster and comparative diagnosis in biomedical problems is the main contribution of this study., The work and the contribution were supported by the SPEV project “Smart Solutions in Ubiquitous Computing Environments”, University of Hradec Kralove, Faculty of Informatics and Management, Czech Republic (under ID: UHK-FIM-SPEV-2019). Supported by the grant TIN2016-75850-R from the Spanish Ministry of Economy and Competitiveness with FEDER funds.

The protection of control and intelligent systems across networks and interconnected components is a significant concern. Biometric systems are smart systems that ensure the safety and protection of the information stored across these systems. A breach of security in a biometric system is a breach in the overall security of data and privacy. Therefore, the advancement in improving the safety of biometric systems forms part of ensuring a robust security system. In this paper, we aimed at strengthening the finger vein classification that is acknowledged to be a fraud-proof unimodal biometric trait. Despite several attempts to enhance finger-vein recognition by researchers, the classification accuracy and performance is still a significant concern in this research. This is due to high dimensionality and invariability associated with finger-vein image features as well as the inability of small training samples to give high accuracy for the finger-vein classifications. We aim to fill this gap by representing the finger vein features in the form of information granules using an interval-based hyperbox granular approach and then apply a dimensionality reduction on these features using principal component analysis (PCA). We further apply a granular classification using an improved granular support vector machine (GSVM) technique based on weighted linear loss function to avoid overfitting and yield better generalization performance and enhance classification accuracy. We named our approach PCA-GSVM. The experimental results show that the classification of finger-vein granular features provides better results when compared with some state-of-the-art biometric techniques used in multimodal biometric systems.

Image processing (IP) is a method of converting an image to digital form by performing operations on it to obtain an improved image or to extract useful information from it. It is a type of signal distribution in which the input is an image such as a video image or a photograph, and the output can be an image or associated characteristics. Besides, this system includes the processing of images in the form of two-dimensional signals, while applying signal processing methods already defined for them. One of the essential steps in IP is a combined application of erosion and dilation procedures, which is part of the mathematical morphology. This article presents a novel thermal image classification based on techniques derived from mathematical morphology. In the processing of grayscale breast cancer images, this method reveals the region of interest as the whitest area.

An image is worth a thousand words; hence, a face image illustrates extensive details about the specification, gender, age, and emotional states of mind. Facial expressions play an important role in community-based interactions and are often used in the behavioral analysis of emotions. Recognition of automatic facial expressions from a facial image is a challenging task in the computer vision community and admits a large set of applications, such as driver safety, human–computer interactions, health care, behavioral science, video conferencing, cognitive science, and others. In this work, a deep-learning-based scheme is proposed for identifying the facial expression of a person. The proposed method consists of two parts. The former one finds out local features from face images using a local gravitational force descriptor, while, in the latter part, the descriptor is fed into a novel deep convolution neural network (DCNN) model. The proposed DCNN has two branches. The first branch explores geo-metric features, such as edges, curves, and lines, whereas holistic features are extracted by the second branch. Finally, the score-level fusion technique is adopted to compute the final classifica-tion score. The proposed method along with 25 state-of-the-art methods is implemented on five benchmark available databases, namely, Facial Expression Recognition 2013, Japanese Female Facial Expressions, Extended CohnKanade, Karolinska Directed Emotional Faces, and Real-world Affective Faces. The data-bases consist of seven basic emotions: neutral, happiness, anger, sadness, fear, disgust, and surprise. The proposed method is compared with existing approaches using four evaluation metrics, namely, accuracy, precision, recall, and f1-score. The obtained results demonstrate that the proposed method outperforms all state-of-the-art methods on all the databases. This work was supported in part by the project “Prediction of Diseases Through Computer Assisted Diagnosis System Using Images Captured by Minimally Invasive and Noninvasive Modalities,” Computer Science and Engineering, PDPM Indian Institute of Information Technology, Design and Manufacturing Jabalpur, Jabalpur, India, under Grant ID: SPARC-MHRD-231, in part by the project IT4Neuro(degeneration) under Grant CZ.02.1.01/0.0/0.0/18 069/0010054, in part by the project “Smart Solutions in Ubiquitous Computing Environments,” Grant Agency of Excellence, University of Hradec Kralove, Faculty of Informatics and Management, Czech Republic under Grant ID: UHK-FIM-GE-2020, in part by the project at Universiti Teknologi Malaysia (UTM) under Research University Grant Vot-20H04, in part by the Malaysia Research University Network (MRUN) under Grant Vot 4L876, and in part by the Fundamental Research Grant Scheme (FRGS) under the Ministry of Education Malaysia under Grant Vot5F073.

Super-resolution is one of the frequently investigated methods of image processing. The quality of the results is a constant problem in the methods used to obtain high resolution images. Interpolation-based methods have blurry output problems, while non-interpolation methods require a lot of training data and high computing power. In this paper, we present a supervised generative adversarial network system that accurately generates high resolution images from a low resolution input while maintaining pathological invariance. The proposed solution is optimized for small sets of input data. Compared to existing models, our network also provides faster learning. Another advantage of our approach is its versatility for various types of medical imaging methods. We used peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) as the output image quality evaluation method. The results of our test show an improvement of 5.76% compared to optimizer Adam used in the original paper [10]. For faster training of the neural network model, calculations on the graphic card with the CUDA architecture were used.

Emerging new technologies and new threads entail additional security; therefore, biometric authentication is the key to protecting the passwords by classifying unique biometric features. One of the protocols recently proposed is the rhythm-based authentication dealing with the dominant frequency components of the keystroke signals. However, frequency component itself is not enough to understand the whole keystroke sequence; therefore, the characteristic of a password could only be analyzed by transformations providing the information of when which dominant frequency arises. In this paper, the biometric signal generated from keystroke data is divided into windows by various window functions and sizes for frequency and time localization. As a guide for signal processing in biometrics and biomedicine, we compared Hamming and Blackman widow functions with various sizes in short time Fourier transformations of the signals and found out that Blackman is more appropriate for biometric signal processing.

In clustering, similarity measure has been one of the major factors for discovering the natural grouping of a given dataset by identifying hidden patterns. To determine a suitable similarity measure is an open problem in clustering analysis for several years. The purpose of this study is to make known a divergence based similarity measure. The notion of the proposed similarity measure is derived from Jeffrey-divergence. Various features of the proposed similarity measure are explained. Afterwards we develop fuzzy c-means (FCM) by making use of the proposed similarity measure, which guarantees to converge to local minima. The various characteristics of the modified FCM algorithm are also addressed. Some well known real-world and synthetic datasets are considered for the experiments. In addition to that two remote sensing image datasets are also adopted in this work to illustrate the effectiveness of the proposed FCM over some existing methods. All the obtained results demonstrate that FCM with divergence based proposed similarity measure outperforms three latest FCM algorithms.

Medical image processing is a highly challenging research area, thus medical imaging techniques are used to make diagnosis in human body. Moreover, as tumor in the brain is a critical and medical complaint, segmentation of the images has an important role to make segmentation of the brain tumor and it provides suspicious region diagnosis from the medical images. By the help of MRI scanners, signals generated by the human body tissues could be detected and determined spatially. Thus, we in this paper try to propose basics of MRI image modalities as a guide for understanding the processes and methods. Since original brain image is not appropriate for the examination, segmentation of the images could be very useful method for partition of the digital image into similar regions. This research also presents a guide for understanding the brain MRI sequences in other words modalities.

Jakub Jirka,1 Michal Prauzek,1 Ondrej Krejcar,2 Kamil Kuca2,3 1Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB – Technical University of Ostrava, Ostrava Poruba, Czech Republic; 2Center for Basic and Applied Research, Faculty of Informatics and Management, University of Hradec Kralove, Hradec Kralove, Czech Republic; 3Biomedical Research Center, University Hospital Hradec Kralove, Hradec Kralove, Czech Republic Objective: The most important part of signal processing for classification is feature extraction as a mapping from original input electroencephalographic (EEG) data space to new features space with the biggest class separability value. Features are not only the most important, but also the most difficult task from the classification process as they define input data and classification quality. An ideal set of features would make the classification problem trivial. This article presents novel methods of feature extraction processing and automatic epilepsy seizure classification combining machine learning methods with genetic evolution algorithms.Methods: Classification is performed on EEG data that represent electric brain activity. At first, the signal is preprocessed with digital filtration and adaptive segmentation using fractal dimensions as the only segmentation measure. In the next step, a novel method using genetic programming (GP) combined with support vector machine (SVM) confusion matrix as fitness function weight is used to extract feature vectors compressed into lower dimension space and classify the final result into ictal or interictal epochs.Results: The final application of GP–SVM method improves the discriminatory performance of a classifier by reducing feature dimensionality at the same time. Members of the GP tree structure represent the features themselves and their number is automatically decided by the compression function introduced in this paper. This novel method improves the overall performance of the SVM classification by dramatically reducing the size of input feature vector.Conclusion: According to results, the accuracy of this algorithm is very high and comparable, or even superior to other automatic detection algorithms. In combination with the great efficiency, this algorithm can be used in real-time epilepsy detection applications. From the results of the algorithm’s classification, we can observe high sensitivity, specificity results, except for the Generalized Tonic Clonic Seizure (GTCS). As the next step, the optimization of the compression stage and final SVM evaluation stage is in place. More data need to be obtained on GTCS to improve the overall classification score for GTCS. Keywords: genetic programming, adaptive segmentation, SVM, fractal dimensions, EEG

Wireless capsule endoscopy (WCE) is one of the most efficient methods for the examination of gastrointestinal tracts. Computer-aided intelligent diagnostic tools alleviate the challenges faced during manual inspection of long WCE videos. Several approaches have been proposed in the literature for the automatic detection and localization of anomalies in WCE images. Some of them focus on specific anomalies such as bleeding, polyp, lesion, etc. However, relatively fewer generic methods have been proposed to detect all those common anomalies simultaneously. In this paper, a deep convolutional neural network (CNN) based model ‘WCENet’ is proposed for anomaly detection and localization in WCE images. The model works in two phases. In the first phase, a simple and efficient attention-based CNN classifies an image into one of the four categories: polyp, vascular, inflammatory, or normal. If the image is classified in one of the abnormal categories, it is processed in the second phase for the anomaly localization. Fusion of Grad-CAM++ and a custom SegNet is used for anomalous region segmentation in the abnormal image. WCENet classifier attains accuracy and area under receiver operating characteristic of 98% and 99%. The WCENet segmentation model obtains a frequency weighted intersection over union of 81%, and an average dice score of 56% on the KID dataset. WCENet outperforms nine different state-of-the-art conventional machine learning and deep learning models on the KID dataset. The proposed model demonstrates potential for clinical applications.