Network Programming
Network programming focuses on designing and implementing algorithms and architectures for processing information across interconnected nodes, aiming to efficiently solve complex computational problems. Current research emphasizes developing novel network architectures, such as graph neural networks and deep operator networks, and improving existing algorithms through techniques like frequency domain inference and tensor decomposition for faster and more accurate computations. These advancements are significant for diverse applications, including improved recommendation systems, enhanced anomaly detection in network flows, and more accurate causal inference from network data. The field's impact spans various scientific disciplines and practical applications, driving progress in areas like machine learning, signal processing, and social network analysis.
Papers
Masked Modeling Duo: Learning Representations by Encouraging Both Networks to Model the Input
Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
Network Aware Compute and Memory Allocation in Optically Composable Data Centres with Deep Reinforcement Learning and Graph Neural Networks
Zacharaya Shabka, Georgios Zervas