Sin Yong Tan

Papers

Total Citations

H-Index

About

Sin Yong Tan is a researcher at the forefront of decentralized machine learning, with a primary focus on developing scalable and efficient algorithms for distributed deep learning systems. His most cited work, "Decentralized Deep Learning Using Momentum-Accelerated Consensus" (2021), introduces a novel framework that enables multiple agents to collaboratively train models without relying on a central parameter server—a critical advancement for privacy-preserving and communication-efficient AI. By integrating momentum-based optimization with consensus protocols, Tan’s approach significantly accelerates convergence in peer-to-peer networks, addressing key bottlenecks in bandwidth-constrained or fault-tolerant environments. While his citation count is currently modest, the work’s foundational nature has already influenced emerging research in federated learning and multi-agent systems. Tan’s contributions are particularly notable for bridging theoretical optimization guarantees with practical decentralized topologies, offering a blueprint for next-generation collaborative AI. His research holds promise for applications ranging from edge computing to autonomous systems, where data locality and robustness are paramount. As the field of decentralized learning matures, Tan’s momentum-accelerated consensus method stands as a pivotal step toward truly distributed intelligence.