Marc Sobel

Papers

Total Citations

H-Index

About

Marc Sobel is a researcher whose work bridges robotics, statistical learning, and computational geometry. His key contributions lie in multi-robot systems, particularly in map merging—a critical challenge for cooperative exploration. In his highly cited 2008 paper (19 citations), Sobel introduced an elegant solution that reduces map merging to the problem of simultaneous localization and mapping (SLAM), enabling multiple robots to efficiently combine local maps into a coherent global representation. This work has practical implications for autonomous navigation and disaster response. In 2006, Sobel advanced statistical methodology with a novel Expectation-Maximization (EM) algorithm derived from the Kullback-Leibler divergence (15 citations). This framework uniquely optimizes both model parameters and the number of components, leveraging nonparametric density estimation to enhance parametric models. It offers a powerful tool for clustering and mixture modeling, impacting fields from bioinformatics to image analysis. Sobel’s additional work on polygonal approximation of point sets (2006) demonstrates his versatility in computational geometry. With over 36 citations across his key papers, his research continues to influence robotics and machine learning, offering practical solutions to complex problems in autonomous systems and data analysis.