863 papers
This paper proposes a state-aware multimodal recommendation ranking model for cross-border e-commerce that explicitly models short-term user intention shifts across six operational states, achieving superior performance in cross-category and low-frequency item recommendations on the newly introduced 42M-record CEMR-42M dataset compared to existing baselines.
This study introduces the MDQ-3L framework to benchmark classical, deep, and quantum machine learning models for air pollution prediction, revealing that while classical models like CatBoost and DBN currently outperform quantum approaches under NISQ constraints, the proposed multi-metric scoring system effectively identifies structural limitations and guides the selection of low-entanglement quantum circuits for future hybrid environmental modeling.
This paper proposes using a low-resource CUT-GAN to generate synthetic mammography patches for training deep learning models in resource-constrained settings, demonstrating that while synthetic data improves model performance compared to no data, it currently yields lower accuracy (0.751 AUC) than training on real images (0.913 AUC).
This paper presents a synthesizable five-stage RV32IM pipeline featuring an integrated gshare branch predictor and a host-independent on-FPGA VGA visualization unit, which achieves a 2.33× performance improvement over a static baseline while enabling real-time pipeline introspection and bit-exact functional verification.
This paper proposes a lightweight Deep Spatio-Temporal Feature Extraction-based Dual-Attention EEGNet (CSFE-DAE) framework that combines Common Spatial Patterns with dual attention mechanisms to achieve state-of-the-art Parkinson's disease detection accuracy of 97.59% on EEG signals.
This study utilized a Four-D development model to create and validate an AI-based learning management system for primary schools, which experts deemed highly feasible and effective in enhancing students' computational thinking skills, particularly in decomposition, abstraction, and algorithms.
This paper proposes a Feature-Adaptive Fusion Hybrid Quantum-Classical Neural Network (FAF-HQNN) that dynamically weights classical and quantum predictions to achieve superior accuracy and robustness against noise in biomedical image classification tasks.
The paper proposes ACDGRec, a robust recommendation model that integrates adversarial training and contrastive learning into a diffusion-based graph neural network to effectively address data sparsity, cold-start issues, and input instability by capturing latent user context correlations.
This paper proposes an NDN-aware cooperative Multi-Agent Deep Deterministic Policy Gradient (MADDPG) framework that leverages decentralized learning and specialized state representations to effectively mitigate Cache Pollution Attacks in Named Data Networking, demonstrating superior cache efficiency, faster convergence, and enhanced resilience compared to single-agent and conventional approaches.
This paper proposes a Linear Mapping-Enhanced Proximal Policy Optimization (LME-PPO) framework that leverages deep reinforcement learning to optimize dynamic load balancing and reduce system latency in smart city edge computing environments by effectively mapping complex task spaces and ensuring computational equilibrium across edge servers.