The rapid development of android applications has been accompanied by a consistent outpouring in sophisticated malware variations that progressively elude signature-based and traditional machine learning detection systems. While considering deep learning models like convolutional and recurrent neural networks, improved results are attained but they rely on local feature extraction and lack in modelling long-range dependencies which limit their adaptability towards advanced obfuscated and polymorphic malwares. In this research work, a lightweight hybrid malware detection framework is proposed which combines MobileNetV2 and a Vision Transformer (ViT) to empower effective and strong Android malware classification from image-based depictions of application binaries. Compact spatial features are extracted by MobileNetV2 and global contextual relationships are captured using self-attention mechanisms by Vision Transformer. Thus, this hybrid design attains effective balance between computational efficiency and detection accuracy. By leveraging the large-scale MalNet dataset comprised of 8 android malware families, the proposed model is evaluated and experimental results presents 98% accuracy, precision, recall, and F1-score by outperforming recent proposed methods based on CNN, transformer, and hybrid baselines. It proves our proposed model maintains compact model size and low inference latency. Further, the results demonstrate that this integration will provide appropriate deployment solution for android malware detection especially real-time security operations which are resource constrained.
Pixel-domain nonconvex priors enhance edge preservation but have limited capability for piecewise planar (PWP) signal reconstruction, whereas gradient-domain convex priors facilitate PWP signal reconstruction but tend to over-smooth fine details. We propose a non-convex graph total variation term in the gradient domain, which promotes PWP signal reconstruction while preserving edges. By combining it with an ℓ 2 -norm fidelity term, we formulate the overall optimization objective. We design an ADMM algorithm with two auxiliary variables and unroll it into a learnable feed-forward network. We further incorporate convolutional attention into the graph learning module to guide gradient weight learning without significantly increasing the number of parameters. Experiments demonstrate that, in blind image denoising and image interpolation tasks, our method outperforms unrolled methods based on pixel-domain nonconvex priors and gradient-domain convex priors, while remaining competitive with general deep learning networks using only a small number of parameters.
In this work, we extend the concept of robust regressions via repeatedly taking the median to nonlinear functions. In the presented method, two nested medians – the Siegel slope – are taken of the n th data derivative for each evaluated polynomial order n , beginning at the highest one. Lower-order coefficients are determined consecutively, by correcting the data for the previously determined ones. We demonstrate and quantify that this algorithm has a high robustness against outliers, with a breakdown characteristic that matches the original linear implementation. We also quantify and rationalize that varying polynomial orders and noise influence the algorithm’s accuracy. The presented method exhibited an increased accuracy and robustness over a wide parameter field, when compared to those of an implementation of the Levenberg–Marquardt algorithm, as well as the method proposed by Theil, namely applying consecutive single median regressions.
Efficient and accurate simulation is essential for the modeling and control of high-frequency DC-DC converters. Conventional numerical integrators require very small time steps to capture fast switching dynamics, leading to high computational cost. This paper introduces a discrete-time analytical integration method that significantly accelerates simulation without compromising accuracy. The approach computes only two or three breaking points per switching cycle, depending on whether the system operates in continuous (CCM) or discontinuous conduction mode (DCM). Each segment of the dynamic response is represented by an exact analytical expression, eliminating the need for fine-grained time-stepping. The method is validated through simulations in both open-loop and closed-loop control scenarios. It also accurately detects the DCM regime and applies rapid corrections to ensure fidelity. The proposed algorithm offers a clear advantage over traditional solvers, enabling fast and reliable simulations suitable for iterative design and real-time applications. It is particularly well-suited for embedded and autonomous systems, such as photovoltaic power units in remote or off-grid settings, and can be extended to more complex converter topologies.
To address low accuracy, difficult parameter determination, and inefficient data use in safety risk prediction, this study proposes an intelligent prediction framework. It uses rough set theory to preprocess data (eliminate incompleteness, discretize continuous data, remove redundancy) and introduces an improved particle swarm optimization (IPSO) by introducing a mutation operator to optimize key parameters of the least squares support vector machine (LSSVM), forming the IPSO-LSSVM model. Experiments on 43 risk samples (33 for training, 10 for testing) show that it outperforms the BP neural network and standard PSO-LSSVM with faster convergence. It enriches integrated intelligent algorithms in safety risk prediction and provides technical references for improving prediction accuracy.
