Cardiovascular diseases are a leading cause of death worldwide, so accurate and timely diagnosis is essential for effective treatment. Electrocardiogram (ECG) signals offer valuable information for detecting various cardiac abnormalities. Traditional ECG analysis is time-consuming and prone to human error, particularly in distinguishing between different types of arrhythmias. Manual interpretation of large amounts of ECG data can lead to misdiagnosis or delayed treatment. There is an urgent need for a reliable solution capable of automatically classifying recorded heart rhythms with high accuracy. In this paper, a deep learning-based framework is presented for classifying ECG heartbeats into five distinct classes. The dataset is preprocessed through normalization and reconstructed into a format compatible with convolutional neural networks (CNNs), which are highly adept at learning spatial hierarchies from input data. The model is trained using optimized hyperparameters and evaluated based on several metrics. The trained model achieves a validation accuracy of 99% and test accuracy of 98% on the ECG Heartbeat Categorization Dataset. The proposed system significantly contributes to the advancement of automated heartbeat classification and has the potential to support clinical decision-making systems.
This paper introduces an improved Data Encryption Standard (DES) framework that incorporates a metrics-driven data processing pipeline, AI-powered analytical decision support and a quantum-inspired entropy-based key generation process to enhance both encryption efficiency and key security. It validates data, preprocesses it, encodes it, optimizes it, encrypts it, monitors it in real-time, analyzes entropy and also analyzes the execution time, encryption speed, throughput, CPU utilization, memory consumption and key randomness. Experimental results indicate that the encryption throughput was achieved on average at 5.11 MB/s, with a maximum value of 12.15 MB/s, while the time required for encryption and decryption were on average 42.67 ms and 44.21 ms, respectively. The proposed quantum-inspired key generator has an average key entropy of 0.82 (compared with 0.68 for standard random keys) and a higher uniformity of 0.88 (compared with 0.72 for standard random keys), which corresponds to an approximately 20.6% improvement. Furthermore, the quantum-inspired approach achieved a maximum entropy of 0.95 compared with 0.89 for random generation, and its strongest keys had a collision rate of no more than 0.01%. The overall average security improvement for the AI analysis was 13.30% and the overall completion rate for the full processing pipeline was 98.50%. The findings highlight practical solutions for optimizing the performance, randomness, and security evaluation of DES-based encryption by incorporating AI-driven analysis, entropy-driven key generation, and regular monitoring.
Federated Learning, often known as FL, is an approach that has recently emerged as a potentially helpful method for training machine learning models in a distributed manner without the requirement of central data storage. However, when attempting to aggregate information, the inherent variety and discrepancies in the data contributed by many FL contributors might be a substantial obstacle. In order to address this problem, researchers have offered various solutions, one of which is called knowledge distillation (KD). Such a solution seeks to transfer knowledge from a larger, more precise model to a smaller model, thus enhancing its performance. This study provides a detailed examination of the effectiveness of KD in responding to these challenges posed by FL. We comprehensively review existing research, emphasizing the benefits and limitations of using these techniques in FL and discussing the numerous challenges and research questions in this field.
Lebanon suffers from financial crises, especially in the electrical sector. Citizens are facing daily power cuts of several hours, leaving many reliant on their own generators or private neighborhood supplies that charge high fees. Fortunately, this country is precious with an abundance of sun, wind, and water. Thus, using renewable energy to generate electricity is considered pivotal for reducing dependency on fossil fuels. Moreover, implementing the concept of energy efficiency through smart home energy management systems is quite promising. Hence, this paper focuses on generating electricity for a smart home using an Adaptive Hybrid Energy System (AHES) consisting of two sources of renewable energies that are available in most Lebanese areas: the sun using solar panels and the wind retrieved from wind turbines, in addition to the batteries and the grid. It will be proven that the suggested system has environmental benefits besides assuring power reliability. It will reduce electricity bills, protect Lebanon from the impact of changing trends in energy markets, reduce the dependency on fossil fuels, and reduce pollution. For testing purposes, an IoT-based implementation of the proposed AHES will be explored, featuring real-time smart home model monitoring using sensors, actuators, and Arduino devices communicating with a cloud server via the Message Queuing Telemetry Transport (MQTT) protocol and a graphical user interface.
