The Probability That a Numerical, Analysis Problem Is Difficult (Classic Reprint) - Softcover

James W. Demmel

 
9781330257166: The Probability That a Numerical, Analysis Problem Is Difficult (Classic Reprint)

Synopsis

Discover how likely numerical problems are to be hard—and why some problems stay solvable.

This book explains a probabilistic way to study numerical conditioning, showing how the distance to ill-posed problems affects accuracy and behavior in algorithms. It ties geometry, probability, and computer arithmetic into a practical view of when and why numerical methods work.

  • Learn the idea that a problem’s condition number connects to how close it is to infinite sensitivity, and how this guides algorithmic reliability.
  • See how uniform distributions and tubular neighborhoods help bound the chances of large errors in matrix inversion, polynomial zero finding, and eigenvalue computations.
  • Understand how finite precision arithmetic changes the picture, and what this means for real-world computations and algorithm design.
  • Explore the limits of the probabilistic model, including when discretization or clustering near ill-posed sets alters expected outcomes.
Ideal for readers who want a clear, theory-grounded view of numerical analysis, conditioning, and the practical behavior of algorithms in computing.

"synopsis" may belong to another edition of this title.

Other Popular Editions of the Same Title