Insider threats remain a serious anxiety for organizations, government agencies, and businesses. Normally, the most hazardous cyber attacks are formed by trusted insiders and not by malicious outsiders. The malicious behaviors resulting from unplanned or planned mishandling of resources, data, networks, and systems of an organization constitute an insider threat. The unsupervised behavioral anomaly detection methods are mostly developed by the traditional machine learning methods for identifying unusual or anomalous variations in user behavior. The insider threat mainly originates from an individual inside the organization who is a current or former employee who has access to sensitive information about the organization. For achieving an improvement over traditional methods, the Stacked Convolutional Neural Network- Attentional Bi-directional Gated Recurrent Unit model is proposed in this paper to detect insider threats. The CNN-Attentional BiGRU model utilizes the user activity logs and user information for time-series classification. Using the log files, the temporal data representations, and weekly and daily numerical features from various sub-models of CNN are learned by the stacked generalization. Based on the chosen feature vectors, a model is trained on the CERT insider threat dataset. The stacked CNN is combined with the Attentional BiGRU model to incorporate more complex features of the user activity logs and user data during each convolution operation without raising network parameters. Thus the classification performance is improved with less complexity. The non-linear time control, chaos-based strategy, update rules, and opposite-based learning strategies are evaluated for generating the Modified-Equilibrium Optimization. The simulation outputs obtained by the model are 92.52% accuracy, 98% Precision, 95% Recall, and 96% F1-score. Thus, the proposed model has reached higher detection performance.

Worldwide, CVDs continue to stand as a prominent contributor to mortality. Detecting and assessing CVD risk early is paramount for effective prevention and management. Medical imaging has gained prominence as a tool for evaluating CVD risk factors. In this study, a novel Inception v3 with a VGG16 is proposed to forecast cardiovascular risk rates via readily available and non-invasive fundus images. This approach harnesses advanced image analysis techniques encompassing contrast enhancement and noise reduction. The blood vessel segmentation and Optic disc detection of the pertinent features are extracted from the fundus images. In this context, the Inception v3 architecture is initially employed to capture intricate hierarchical patterns within the images. Alternatively, explore the utilization of the VGG16 architecture. By integrating these features with clinical data, the model is then trained to predict cardiovascular risk rates. Empirical findings underscore the method's remarkable accuracy in risk rate prediction. This non-invasive, image-based methodology holds transformative potential for reshaping early diagnosis and risk management approaches for cardiovascular diseases. Ultimately, this innovation stands to enhance patient care and outcomes.

A large number of association rules often minimizes the reliability of data mining results; hence, a dimensionality reduction technique is crucial for data analysis. When analyzing massive datasets, existing models take more time to scan the entire database because they discover unnecessary items and transactions that are not necessary for data analysis. For this purpose, the Fuzzy Rough Set-based Horse Herd Optimization (FRS-HHO) algorithm is proposed to be integrated with the Map Reduce algorithm to minimize query retrieval time and improve performance. The HHO algorithm minimizes the number of unnecessary items and transactions with minimal support value from the dataset to maximize fitness based on multiple objectives such as support, confidence, interestingness, and lift to evaluate the quality of association rules. The feature value of each item in the population is obtained by a Map Reduce-based fitness function to generate optimal frequent itemsets with minimum time. The Horse Herd Optimization (HHO) is employed to solve the high-dimensional optimization problems. The proposed FRS-HHO approach takes less time to execute for dimensions and has a space complexity of 38

Wireless Sensor Network (WSN) communication encounters security vulnerabilities, particularly with network traffic being susceptible to attacks during routing. The effective use of Deep Learning (DL) methods has been demonstrated in developing Intrusion Detection Systems (IDSs) to manage security attacks in Wireless Sensor Networks (WSN). Consequently, the development of new IDS becomes imperative, with DL and optimization algorithms offering superior attack detection capabilities. To address this need, we propose one new IDS by integrating Fuzzy Temporal rules and Artificial Bee Colony (ABC) optimization algorithm with Convolutional Neural Network (CNN) optimized with (FT-ABC-CNN) to enhance the classifier performance. To assess its effectiveness, a comparative analysis was conducted between the newly proposed FT-ABC-CNN algorithm and other classification algorithms commonly employed in Intrusion Detection System design, such as CNN, Long Short-Term Memory (LSTM), and Recurrent Neural Networks (RNN). Experimental evaluations revealed that the FT-ABC-CNN algorithm surpassed these comparable classifiers in terms of accuracy enhancement and reduction in false positive rates.

In the evolving realm of Mobile Edge Computing (MEC), efficient task offloading remains pivotal. This paper introduces the Hybrid Energy-Efficient Task Offloading Algorithm (HEETA) to address the deficiencies of the current Joint Optimization Task Offloading Strategy based on Particle Swarm Optimization (JOBPSO). Drawing from a broad dataset encompassing diverse MEC operational variables, HEETA exhibits exemplary performance metrics, with a notable mean fitness of approximately 0.99984 and a minimal standard deviation of 0.000116. Such metrics not only reflect HEETA's robustness but also its adaptability across multifaceted MEC parameters. Furthermore, its dynamic nature facilitates adaptability to variables including task numbers, computational capacity, and latency constraints, resulting in marked improvements in energy efficiency. Quantitative evaluations, as evidenced by a performance matrix, position HEETA's global best fitness values between approximately 0.9998 and 0.9999. While HEETA signifies a monumental step in enhancing energy efficiency, prolonging device longevity, and optimizing overall MEC system performance, the research acknowledges potential limitations, emphasizing the imperatives of accurate modeling and subsequent validations within distinct MEC environments.

Secure routing and communication with confidentiality based on encryption of texts in multiple natural languages are challenging issues in wireless sensor networks which are widely used in recent applications. The existing works on Elliptic Curve Cryptography based secured routing algorithms are focused only on the encryption and decryption of single language text encrypted over a Prime finite field. In this article, a new algorithm called Multi-Language ECC encrypted Secure Routing algorithm with trust management is proposed, in order to ensure confidentiality and integrity which focuses on the encryption of plain text using Riemann's zeta function and Elliptic Curve Cryptography for improving the key strength which is applied for encryption over a range of multi languages namely Tamil, English, Hindi French and German which are supported by Unicode and routing the text security. From the experiments conducted using the proposed multi-lingual encryption algorithm with network routing, we prove that the suggested method provides greater security than the current secure routing algorithms due to the use of Zeta function and Gamma function with ECC key and trust management. but also boasts reduced complexity compared to other existing multi-lingual encryption algorithms.

Design of effective algorithm for reliable and energy optimized secure routing protocol (SRP) for wireless sensor networks (WSNs) is a demanding design issue now. To handle this problem, we propose a trust and encryption-based SRP based on trust modelling with intrusion detection, elliptic curve cryptography (ECC), clustering, fuzzy rules and ant colony optimization (ACO)-oriented SRP for WSN routing. In this paper, an extended convolutional neural networks with Schrodinger equation and particle swarm optimization is proposed for developing and intrusion detection-based trust modelling. Moreover, a new node authentication scheme and an encryption-based secure routing protocol are also proposed in this work for increasing the security. This proposed secure protocol known as trust and ECC encryption-based ACO-SRP (TECC-ACO-SRP) performs authentication and trust analysis on the nodes using intrusion detection, and then, the data are communicated after data encryption using ECC encryption technique. This proposed system combines dominant set clustering with fuzzy rules to make clusters with similar type of nodes as members and then selects cluster heads (CHs) for every cluster. This SRP ensures improved security, reduced delay and energy usage with higher packet delivery ratio than other existing SRPs.

Intensive Research in the area of Hardware & Networking involves Security. This key concept is bypassed in the initial Design & Development of Hardware devices (IoT), paving way for vulnerable attacks. Attacks in Hardware Devices (IoT) are broadly classified into (i) Active attacks (ii) Passive attacks. This survey focusses on Passive attacks that include side-channel analysis aimed at retrieving data or scanning open ports and network vulnerabilities. In this Technological era, Side Channel Attacks are threatening the whole world, because communication is almost digital everywhere and anywhere. Security experts and cryptologists are working on a method to prevent such attacks on cryptographic implementations or devices, in order to prevent personal information from being leaked or misused. Adapting ML/DL methodologies for providing security has been advocated in a number of researches. This paper gives an overall survey of all the Side channel attacks involving the execution of Cryptanalytic algorithms and its counter measures encountered in the last few decades.

In Wireless Sensor Networks (WSNs), sensor nodes are placed to sense and collect data. Due to the energy constraint nature of WSN, optimising the energy during the data dissemination is a major concern. To solve this problem, data aggregation may be used to bring down the redundant transmission of packets in WSN. In most of the previously available techniques, security is also a major concern during data aggregation and routing process with optimized energy. Many data aggregation-based routing systems are subject to security attacks during the data transfer from sensors to source to Clustered Heads (CHs) and then to sink with data aggregation process. Moreover, the existing data aggregation-based routing protocols suffer from data redundancy with less accuracy in aggregated data. For handling such issues order to overcome these issues, an Energy Efficient Secured Clustered Particle Swarm Optimization (PSO) oriented Data Aggregation Routing Protocol (EESCPSO-DARP) that can provide efficient authentication during data aggregation-based routing is introduced in this paper. Moreover, the proposed protocol enhances the rate of the data transmission by efficient prevention of false data injection and other attacks through node authentication and data encryption. This proposed protocol minimizes the energy usage by minimizing the retransmissions by eliminating the possible redundant transmissions of data during data aggregation-based routing. Moreover, the introduced protocol minimizes both communication and computational overhead through optimal clustering and routing with PSO and provides and efficient routing system. This proposed EESCPSO-DARP protocol has been developed by using the NS3 simulator. The results of this protocol showed improved security, higher packet delivery ratio and enhanced network throughput with reduced energy and delay.

Everyone likes to be attractive. Many people gain self-confidence in the presence of other people. Spectacles are arguably one of the most significant instruments to display one's attitude, in addition to any stylish accessories. Choosing the best frames is the most difficult struggle for a person. Recommending glasses based on their face shape is the main objective of this project. This project works on the automatic extraction of face shapes and classification methods for spectacles frame shapes. Convolutional Neural Network technology was used in the development of this system to choose the best eyewear based on the user's facial shape. The person's face is taken as an input image with good accuracy and the best frame for their face shape is recommended. This research will apply the concepts and principles of an innovation that can provide the frame shape of the glasses.

High-Efficiency Video Coding (HEVC) has a higher coding efficiency, its encoding performance must be increased to keep up with the expanding number of multimedia applications. Therefore, this paper proposes a novel Rectified Linear Unit-Bidirectional Long Short-Term Memory-based Tree Social Relations Optimization (ReLU-BiLSTM-based TSRO) method to enhance the quality of video transmission. The significant objective of our proposed method aims in enhancing the standards of entropy encoding process in HEVC. Here, context-adaptive binary arithmetic coding (CABAC) framework which is prevalent and an improved form of entropy coding model is utilized in HEVC standards. In addition to this, the performances of the proposed method are determined by evaluating various measures such as mean square error, cumulative distribution factor, compression ratio, peak signal-to-noise ratio (PSNR) and bit error rate. Finally, the proposed method is examined with five different sequences of video from football, tennis, garden, mobile and coastguard. The performances of the proposed method are compared with various approaches, and the result analysis shows that the proposed method attained minimum mean square error (MSE) loss with maximum PSNR rate.

Information Technology systems are more and more susceptible to many variety of security risks, the majority of risks are mostly started by internal users. Insider threats are a complicated and difficult problem to identify and avoid because they mostly have privileged access to the network and database of company details. Businesses, organisations, and governmental organisations have a serious cybersecurity risk as a result of insider attacks. Insider threat identification is challenging because of uneven data, scant ground truth, and potentially changing user behavior. This work provides an insider threat detection method based on anomaly detection using unsupervised learning. Computer network and system security are severely hampered by insider threats. Theft of intellectual property, sabotage, the release of sensitive data, and web application attacks are examples of malevolent actions committed by authorised users that have the potential to do serious harm. It is the responsibility of the organisation to protect all network layers and guard against intrusions. Using historical data, the extracted behavioral traits using the Deep Learning method. The project made advantage of the publicly accessible Computer Emergency Response Team (CERT) insider threats dataset.

The most difficult risks in cyber world are insider threats, because of the modest nature of the insider threat it is quite difficult to identify attackers. Over the last few decades, insider threat detection has risen in popularity. Insider threats are one of the most difficult risks on the internet, and they usually pose serious complications for the organization. Insider attacks are perpetrated by persons who have lawful access to an organization's network, applications, or databases. To analyze the framework execution, many execution methodologies have been identified to work with insider danger situations. Because of the nuanced and adaptive nature of insider threats, heterogeneity, and complexity, it is challenging to identify the behavioral differences between insider users and regular users due to the lack of defined insider risks. On the basis of the audit raw data, network, or ambient data-based the problem of identifying security breaches has been taken into consideration. Then, each piece of work is assessed based on its capacity to defend against insider threats, the manner in which data is gathered from the relevant data sources, and the type of the algorithm used to deal with situations.

Radiologist diagnose the brain disease through shape and boundary regions of brain in medical image such as CT, MRI, and PET. Automatic medical image segmentation and enhancement method perform less in boundary regions due to artefacts such as dense objects and slice overlap. Manual enhancement and segmentation method never differentiates the shape and location of regions in brain CT/MRI images. Dyadic cat optimization (DCO) algorithm is proposed for segmenting brain regions in medical images such as CT and MRI through Nonlinear perspective Foreground and Background projection. DCO algorithm eliminates the artefacts in the boundary regions of brain and enhances the boundaries and shape such as pterygomaxillary fissure, occipital lobe, vaginal process, zygomatic arch, maxilla and piriform aperture for more visibility. Proposed DCO algorithm enhances the occipital lobe and zygomatic arch regions in CT/MRI image. The occipital lobe and zygomatic arch regions are better enhanced and segmented with DCO algorithm than traditional algorithm and achieve an accuracy of 90% through structural similarity index and visual interpretation.

The concept based on data mining has drawn considerable attention from various database professionals and research scholars. The progression of computer-based advancements, namely database management and data storage has facilitated the storage of large data and the data mining approaches are employed to gain valuable information from huge databases. Recently, several techniques to association rule mining (ARM) and frequent itemset mining (FIM) have been established; yet the efficiency based on execution time and scalability continues to be seen as a significant limitation that results in poor solution quality. Therefore, it is necessary to enhance the consistency that signifies the total number of frequently discovered frequent itemsets. This paper proposes three different phases namely the pre-processing phase, FIM phase and ARM phase. In the first pre-processing phase, the Twitter databases are pre-processed and converted into a suitable format for FIM. Here, the tweets are converted into related feature sets and items. In the second FIM phase, an improved Apriori algorithm is 1utilized in mining and extracting the frequent Then in the final phase, an adaptive billiard inspired optimization (ABIO) algorithm which is the integration of neural network (NN) optimization algorithm and billiard inspired optimization (BIO) algorithm is proposed for the optimal generation of association rules with minimum support and confidence from the huge itemsets. Finally, the recent tweets based on covidvaccine, BTSlivestreaming, KFC, McDonald’s as well as lockdown achieved using the hashtag is evaluated for various performance measures, like precision, recall, [Formula: see text]-measure, execution time and memory utilization. Also, comparative analyses are performed to evaluate the efficiency of the proposed technique.

Neurologist analyses shape and structure of brain parts through any medical images such as CT, MRI, and PET for disease diagnosis. For diagnosis, automatic medical image segmentation segments the parts of brain with low contrast, and artefacts are never removed over boundary region in different parts of brain. Manual segmentation shows poor differentiation in boundary regions due to artefacts or steaks. In this paper, we propose dyadic CAT optimisation (DCO) algorithm for segmenting the brain regions from CT and MRI images via nonlinear perspective foreground and background projection. DCO algorithm provides exact structure and shape of brain regions and eliminates artefacts in boundary regions. DCO algorithm delineates the boundary region such as dentate nucleus, pontine tegmentum, pontine nuclei, petrosal nerve, petrous part of temporal bone, crista galli, internal occipital crest, and mastoid emissary foramen in brain image with high visibility and enhanced boundary and differentiates deformable shape. Performance of DCO algorithm is evaluated through 50 MRI and CT brain images and eight images with complex bone and muscle mass structures of brain. DCO algorithm shows an accuracy of 90% through structural similarity index.

Nowadays, the concept of data mining is employed widely and created a great deal of attention due to its fast arrival. Numerous approaches to frequent itemsets and association rule mining (ARM) are exemplified in recent years, but still, the performances based on scalability and processing time are considered as a major drawback that results in obtaining the solutions with very poor quality. To overcome such shortcomings, this article proposes three significant phases, namely, the data pre‐processing phase, data pre‐processing, frequent itemset mining, and ARM. In data pre‐processing phase, the collected twitter datasets are pre‐processed to eliminate redundant data and convert them into an appropriate format for further mining. In the frequent itemset mining phase, an Apriori algorithm is employed for the exact mining of frequent itemsets. The ARM phase utilizes the fuzzy manta ray foraging (FMRF) optimization algorithm that involves the generation of association rules from the huge itemsets thereby achieving minimum confidence and minimum support value. Here, the recent tweets regarding Covid‐19, trump2020, joebiden, draintheswamp, and Godzilla are the datasets collected from the Twitter web link. The experimental analysis and the comparative performances are performed for various simulation measures and the results reveal that the proposed approach provides effective performances when compared with various other existing approaches.

This paper aims to design and implement an application usage behaviour analytics system to produce dynamic bundles personalized to each customer. A H2O Deep Learning model was first built to predict the list of products that are most likely to be reordered by each customer, on the basis of which bundle combinations were framed. The model was found to have an accuracy score of 83.04, a precision score of 84.7 and a recall score of 96.31. The products comprising each customer’s bundle were determined by a 3:1 ratio of products that the customer is most expected to purchase to the products that are less popular. This implementation starts with building a data lineage and catalog from multiple data sources with persistent data collected from transactions and product catalogs which was then used to identify, model, analyze behavioral patterns and collect insights in order to build the prediction model. This method of recommendation attempts to solve the cold-start effect of new products and to eliminate the popularity bias in bundle recommendation systems by producing personalized dynamic bundles in such a way that it improves the sales of unpopular products while also contributing to overall customer satisfaction. The technical stack that was used for implementing this idea includes Apache Spark, Mongodb Atlas and Apache H2O.

The neurologist analyses the brain images to diagnose disease via structure and shape of the part in scanned Medical images such as CT, MRI, and PET. The Medical image segmentation performs less in the regions where no or little contrast, artifacts over the different boundary regions. The manual process of segmentation shows poor boundary differentiation due to discernibility in shape and location, intra and inter observer reliability. In this paper, we propose dyadic CAT optimization (DCO) algorithm to segment the regions in the brain from CT and MRI image via Non-linear perspective Foreground and Back Ground projection. The DCO algorithm removes the artifacts in the boundary regions and provide the exact structure and shape of the brain regions. The DCO algorithm shows the region boundary for pterygomaxillary fissure, occipital lobe, vaginal process zygomatic arch, maxilla and piriform aperture in brain image with high visibility in the regions of inadequately visible boundary and distinguishes the deformable shape. The DCO algorithm applies on 50 images and eight images with complex bone and muscle mass structure for performance evaluation. The DCO algorithm shows the increased Structural similarity index (SSIM) with 90% accuracy.

Numerous public networks, namely Instagram, YouTube, Facebook, Twitter, etc., share their own feelings and idea as videotapes, posts, and pictures. In future research, adapting to such data and mining valuable information from it will be an undeniably troublesome errand. This paper proposes a novel audio–video–textual-based multimodal sentiment analysis approach. The proposed approach investigates the sentiments that are collected from the web recordings that utilize audio, video, and textual modalities for further extraction. A feature-level fusion technique is employed in fusing the extracted features from different modalities. Therefore, the extracted features are optimally chosen by using a novel oppositional grass bee optimization (OGBEE) algorithm to obtain the best optimal feature set. Here, 12 benchmark functions are developed to validate the numerical efficiency and the effectiveness of a novel OGBEE algorithm for various aspects. Moreover, our proposed approach utilizes multilayer perceptron-based neural network (MLP-NN) for sentiment classification. The experimental analysis reveals that the proposed approach provides better classification accuracy of about 95.2% with less computational time.