3020 papers
This paper introduces a lower-tail risk-aware framework for fair influence maximization that utilizes Conditional Value-at-Risk and worst-group shortfall probability to protect disadvantaged groups from severe coverage failures, achieving a significant reduction in coverage risk with only a marginal loss in total expected reach.
This study extends modified wavenumber analysis to Physics-Informed Neural Networks (PINNs) to systematically evaluate their spectral accuracy on high-frequency wave problems, revealing that deeper architectures, specific activation functions like sinusoidal or tanh-Gaussian, and an optimal number of Fourier modes () are critical for minimizing dispersion and dissipation errors.
This paper proposes LARC-QL, a Q-learning enhanced content caching scheme for Content-Centric Networks that utilizes dual Q-tables and real-time demand signals to dynamically optimize caching decisions and on-path placement, significantly improving cache hit ratios, latency, and resource efficiency compared to static strategies.
This paper demonstrates that Quantum Autoencoders, specifically an 8-feature model with Dense-Angle encoding and a RealAmplitude ansatz, outperform Classical Autoencoders in cybersecurity anomaly detection on the BETH dataset by achieving a higher F1 score (0.87 vs. 0.77) even when trained on substantially fewer samples.
This systematic review of 20 peer-reviewed studies evaluates the post-COVID landscape of AI-powered mobile proctoring in higher education, highlighting its potential for scalable exam integrity while critically addressing significant gaps in mobile-specific research, technical reliability, and ethical concerns such as privacy and algorithmic bias.
This paper introduces AllLinLog, a streamlined model that eliminates the need for log parsing by using a linear self-attention encoder to directly process full contextual information from system logs, thereby achieving superior anomaly detection accuracy and faster inference compared to traditional methods.
This paper proposes and evaluates a hybrid cryptographic model that integrates RSA with Quantum Key Distribution (using the BB84 protocol) to secure consumer electronics against quantum computing threats, demonstrating enhanced performance in key generation, encryption speed, and resistance to man-in-the-middle attacks.
The paper proposes DOKA-Agent, a large language model agent framework that employs a perceive-plan-act workflow to achieve dynamic, incremental ontology alignment in evolving knowledge graphs, thereby improving semantic precision and reducing update latency compared to traditional static methods.
Using a cybersecurity-in-space corpus, this study demonstrates that while graph topology offers context-dependent incremental value in forecasting technology emergence, node-level temporal histories already encode the majority of predictive information, resulting in only modest and non-uniform improvements from graph-aware models over strong temporal baselines.
This paper presents CuSPC, a GPU-oriented architecture for steganographic polar coding that restructures decoding schedules and leverages multi-level parallelism to achieve a 30-fold throughput increase over CPU baselines while maintaining low coding loss and comparable steganalysis resistance.