Face Emotion Recognition (FER) is a form of biometrics used to analyze and deduce the emotional state of any individual. Several challenges arise while implementing such systems due to the variability in expression depending on factors like age and culture, variance in head poses restricting the emotion to be captured, and occlusions degrading the quality of images. With the attention now turning towards Deep Learning (DL), Convolutional Neural Networks (CNN) are the predominant method where the parameters can get tuned, i.e., the number of layers to discover the preferred model for the given problem. This study experiments with a CNN model incorporating 12 layers, operating on two datasets to detect various emotions. A comparison is made between the performance of the CNN model using visual metrics, including confusion matrix and classification reports, along with evaluation metrics such as accuracy, precision, recall, and F1-score. The experimented CNN model obtained an accuracy of 86% with Young AffectNet-HQ and 92% achieved with AffectNet-HQ, thereby showing promising results.

Fog computing (FC) is a concept that encompasses cloud paradigm-like amenities to the edge of the network. The fog layer that complements the cloud-based concept aids in strengthening the standards of services (QoS) in time-sensitive activities. One such application of FC arises in the deployment of content delivery networks for reducing the latency in transferring content to customers and enhancing the user experience. This paper presents a fog-based system where fog nodes act as caches for content storage. Appropriate placement of cache nodes is important for fast content distribution to users and maintaining the QoS. Another challenge is to minimize energy expenditure while deploying caches at scale. To deal with these issues, a Connected Dominating Set (CDS) and popularity-based caching strategy in content distribution fog networks are suggested in this work. The content is sorted based on its past popularity and then CDS based rules have been developed to recommend the most ideal fog nodes (FNs) to place popular content or files. The selection of fog nodes as caches is performed based on factors of storage capacity, energy consumption, and user density (i.e., the number of neighbors of a fog node). Experiment outcomes authenticate the efficiency of the anticipated scheme without caching and random method of cache placement. In comparison to the without caching and random method, the proposed method produces better fog nodes as cache nodes, incurs a lower energy consumption and consumes less bandwidth.

Cement is a basic material for all types of construction. This paper deals with the probabilistic assessment of two reliability models of a clinker manufacturing system running in cement plants consisting of three components: crusher, roller mill and rotary kiln. The maintenance and rest are turned on over the span of no demand. Preventive/corrective maintenance and general checking is also done randomly for the system. The failure rate, repair rate and cost for repair/replacement have been calculated from the data referred from Shree Cement Ltd., Khushkhera, Rajasthan, India. The factors of system effectiveness including the profit incurred to the system are obtained using semi-Markov processes and regenerative point technique. Various results useful to cement industries are interpreted graphically. Furthermore, the outcomes of this study can be helpful to different cement plants by upgrading the components’ efficiency and planning for reliability increase.

Machine learning algorithms have substantially increased the ability of intrusion detection (IDS) systems to identify and categorize cyber-attacks on the network and host levels in real time. But the fact that dangerous attacks happen frequently and are always evolving causes a number of issues that call for scalable solutions. This article discusses how to develop adaptive and powerful intrusion detection systems that can identify and classify unwanted and unplanned cyber-attacks using deep neural networks (DNNs), a form of deep learning model. A number of benchmarks damaging datasets that are accessible to the general public are provided with a full examination of DNN and other conventional machine learning classifier trials. The KDDCup99 and NSDL-KDD datasets are used in the following hyper parameter selection technique to determine the best DNN network parameters and network design. System administrators can use Network Intrusion Detection Systems (NIDS) to investigate network security vulnerabilities in their enterprises. To counter unconventional and unplanned assaults, there have been numerous attempts to develop reliable and effective NIDS. Each DNN experiment has 1,000 epochs and a learning rate between [0.01-0.5]. DNNs outperform traditional machine learning classifiers, according to the results of rigorous experimental testing.

Identity and Access management (IAM) [1] plays an important role when it comes to background verification. It is a great way to know people you are working with, whether it is a professional front or some local business. Identity theft and secure document exchange are major issues with the current scenario and blockchain offers to be a great solution. The introduction of the public key, private key, transaction verification and foot printing will play a significant role in securing IAM. The idea is to store user documents and other critical information inside the block chain. All the verification is given based on the user’s consensus to the particular request which will trigger further functionalities, responsible for secure data exchange. According to the property of blockchain, the chain will contain the history of each transaction that keeps track of every user-company activity that will prevent any action which is against the will of both parties.

In the current digital epoch, the vast expanse of information has revolutionized the accessibility of knowledge and perspectives. Nevertheless, this information abundance has introduced challenges in navigating and comprehending the deluge of textual data. The surge in online news articles, research papers, reports, and diverse document genres has accentuated the necessity for proficient document summarization techniques. Traditional manual methods of summarization are time-intensive and influenced by subjective biases. In contrast, the synergy between Natural Language Processing (NLP) and machine learning has unlocked the potential for automated document summarization, promising efficient information consumption and informed decision-making. This research paper delves into the convergence of these factors. It is driven by the Longformer model's distinctive capability to manage extensive texts while retaining contextual coherence—a potential solution to the hurdle of large document summarization. By capitalizing on the Longformer's architecture, this study endeavors to exploit its prowess in generating cohesive summaries from lengthy source documents, thereby amplifying the accessibility of intricate information.

Medical data is incredibly sensitive to privacy issues and comprises much personal information. Medical data must be correctly and securely kept in the era of big data due to the growing digitization of the healthcare industry. However, sharing current health information is challenging and prone to the danger of privacy breaches. This article suggests a healthcare information security storage solution based on the Mandatory Access Control and Usage Control (UCON) architecture and the Corda Blockchain to address these problems. The plan uses Mandatory Access Control (MAC) and Usage Control (UCON), which permits dynamic and granular access to medical data before storing the data on the blockchain, which can be made secure and impenetrable by creating relevant smart contracts. Furthermore, this solution integrates IPFS technology to alleviate the blockchain's storage load. Experiments show that the proposed scheme in this paper, which combines Mandatory Access Control (MAC), Usage Control (UCON), and blockchain technology, not only ensures the secure storage and integrity of medical information but also defines usage policies that specify how sensitive data can be used.