Heart Disease (HD) diagnosis can be imperative by amalgamating IoT and AI, as IoT devices permit continuous, real-time patient data monitoring, when AI algorithms afford precise, personalized risk assessments and extrapolative analysis. This approach is practical because it enables early detection and involvement, thereby enhancing patient outcomes by addressing potential problems before they escalate. Earlier research has frequently been constrained by data collection in static, inadequate personalization and a lack of real-time analysis. To address the challenges in heart disease diagnosis within smart healthcare environments, we propose a novel Double Level Risk-based Personalized Heart Disease Detection Model (DLR-PHDM). Similar to our previous work, this model begins with data acquisition using IoT devices and pre-processing at the Distributed Edge Server (DES). The pre-processed data is then fed into an Attention-based Long Short-Term Memory-Capsule Network (ALSTM-CapsNet) for classifying patient risk into three levels: low, moderate and high, based on Electronic Health Records (EHRs) and current sensor data. The risk classification results are managed by a Distributed Reinforced Scheduler (DRS) for Heart Disease Crisis Planning (HDCP), where patient data is scheduled according to their risk level using Multi-Agent Deep Reinforcement Learning (MA-DRL). This scheduling takes into account factors such as patient risk score, resource availability, wait time and appointment slot utilization. Subsequently, the scheduled reports are sent to the Multi-Cloud Server (MCS) for heart disease detection, utilizing tailored AI models: Support Vector Machine (SVM) for low-risk, Convolutional Neural Network (CNN) for moderate risk and Transformer-based CNN (T-CNN) for high-risk patients. The detected results are then used to provide personalized recommendations to the corresponding patients. During implementation, the proposed DLR-PHDM model demonstrated superior performance, achieving an accuracy of 99.79%, precision of 99.10%, specificity of 98.62%, sensitivity of 99.95% and an F-score of 99.12%. It outperformed the state-of-the-art heart disease prediction models.

The rapid growth and diversification of malware variants, driven by advanced code obfuscation, evasion, and antianalysis techniques, present a significant threat to cybersecurity. The inadequacy of traditional methods in accurately classifying these evolving threats highlights the need for effective and robust malware classification techniques. This article presents WinDroid, a novel visualization‐based framework for Windows and Android malware family (AMF) classification using hybrid features and hierarchical ensemble learning. The WinDroid system employs a multistage approach to malware classification, transforming binaries into Markov grayscale images, enhanced via contrast‐limited‐adaptive‐histogram‐equalization and gamma correction. Deep learning and handcrafted features are extracted and fuzed using graph attention networks (GATs), feeding into hierarchical support vector machines (SVMs) for accurate family classification. This framework effectively reduces information loss, enhances computational efficiency, and demonstrates outstanding performance. WinDroid delivers excellent results, achieving 99.53% accuracy on Windows and 99.65% on AMF classification, along with Cohen’s kappa coefficients of 99.01% and 99.28%, respectively, and outperforming state‐of‐the‐art baseline methods.

Cybercriminals motivated by malign purpose and financial gain are rapidly developing new variants of sophisticated malware using automated tools, and most of these malware target Windows operating systems. This serious threat demands efficient techniques to analyze and detect zero-day, polymorphic and metamorphic malware. This paper introduces two frameworks for Windows malware detection using random forest algorithms. The first scheme uses features obtained from static and dynamic analysis for training, and the second scheme uses features obtained from static, dynamic, malware image analysis, location-sensitive hashing and file format inspections. We carried out an extensive experiment on two feature sets, and the proposed schemes are evaluated using seven standard evaluation metrics. The experiment results demonstrate that the second scheme recognizes unseen malware better than the first scheme and three state-of-the-art works. The findings show that the second scheme’s multi-view feature set contributes to its 99.58% accuracy and lowers false positive rate of 0.54%.