Overview
The International Symposium on Scientific Computing, Computer Arithmetic, and Validated Numerics (commonly abbreviated SCAN) is a biennial meeting that brings together researchers and practitioners working on numerical analysis and reliable computation. The symposium emphasizes both theoretical developments and practical implementations that improve the accuracy, efficiency, and trustworthiness of numerical software and algorithms.
Core topics
Sessions typically cover a range of closely related fields. Representative topics include:
- numerical analysis and error estimation
- numerical integration and quadrature methods
- numerical linear algebra, solvers and preconditioning
- numerical methods for partial differential equations and discretization techniques
- interval arithmetic and its software implementations
- validated numerics for guaranteed error bounds and rigorous computation
History and organization
SCAN evolved as an international forum for topics that combine numerical mathematics with concerns about numerical reliability and hardware/software arithmetic. It is typically organized by a rotating local committee, often associated with a university or research institute, and draws contributions from Europe, Asia, the Americas and elsewhere. Proceedings are usually peer reviewed and published either as a conference volume or as special issues of scholarly journals.
Program, participants and formats
A typical program blends invited plenary talks, contributed sessions, poster sessions and tutorials or workshops on software tools. Attendees include academic researchers, graduate students, and engineers who develop or apply numerical libraries. Demonstrations of interval and certified-computation libraries, examples of validated simulation, and reproducibility discussions are common components.
Significance and distinctions
SCAN is notable for its explicit focus on rigorous and reliable computation rather than only on performance or asymptotic theory. It fosters cross-disciplinary exchange between mathematicians, computer scientists and domain scientists concerned with trustworthy numerical results. For people working on error control, certified algorithms or computer arithmetic design, SCAN provides a concentrated venue to share methods, software and benchmarks.
Further information, past programs and calls for papers are normally available from the symposium organizers or affiliated societies. For typical topic descriptions see the linked subject entries above.