Jonathan Richard Shewchuk

Jonathan Richard Shewchuk is a Professor in the Computer Science Division at the University of California at Berkeley. His research focuses on computational geometry (mesh generation, Delaunay triangulations), scientific computing, numerical methods, and neural network mathematics. He is known for developing Triangle, a widely-used 2D mesh generator, and Stellar, a tool for improving tetrahedral mesh quality. His work on robust geometric predicates and conjugate gradient methods has had lasting impact.

Education:

• B.Sc. in Physics and Computing Science, Simon Fraser University (1990)
• M.S. and Ph.D. in Computer Science, Carnegie Mellon University (1997)

Research Interests: Shewchuk's work combines theoretical rigor with practical software development. He has pioneered methods for handling large-scale geometric data (e.g., LIDAR scans) and contributed foundational algorithms for mesh generation and numerical robustness. His recent focus includes the mathematical foundations of linear neural networks and their weight spaces.

Notable Software:

• Triangle (2D mesh generation)
• Stellar (tetrahedral mesh improvement)
• Streaming GIS tools for massive data processing

Students & Collaborators: Advised PhD students including Martin Isenburg (LAStools developer) and Marc Khoury. Collaborated with Jack Snoeyink, Yuanxin Liu, and others on streaming geometry algorithms.

Labs/Teams:曾参与伯克利计算机动画与建模组 (Berkeley Computer Animation & Modeling Group), focusing on physically-based animation and dynamic geometry modeling.