Universality in numerical computations with random data
Authors:
Open Access
- 24 September 2014
- journal article
- research article
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 111 (42) , 14973-14978
Abstract
The authors present evidence for universality in numerical computations with random data. Given a (possibly stochastic) numerical algorithm with random input data, the time (or number of iterations) to convergence (within a given tolerance) is a random variable, called the halting time. Two-component universality is observed for the fluctuations of the halting time—i.e., the histogram for the halting times, centered by the sample average and scaled by the sample variance, collapses to a universal curve, independent of the input data distribution, as the dimension increases. Thus, up to two components—the sample average and the sample variance—the statistics for the halting time are universally prescribed. The case studies include six standard numerical algorithms as well as a model of neural computation and decision-making. A link to relevant software is provided for readers who would like to do computations of their own.Keywords
All Related Versions
This publication has 19 references indexed in Scilit:
- A neural computation model for decision-making timesJournal of Mathematical Psychology, 2012
- Universality of local spectral statistics of random matricesBulletin of the American Mathematical Society, 2012
- Jacobi’s Method is More Accurate than QRSIAM Journal on Matrix Analysis and Applications, 1992
- Eigenvalues and Condition Numbers of Random MatricesSIAM Journal on Matrix Analysis and Applications, 1988
- The probability that a numerical analysis problem is difficultMathematics of Computation, 1988
- GMRES: A Generalized Minimal Residual Algorithm for Solving Nonsymmetric Linear SystemsSIAM Journal on Scientific and Statistical Computing, 1986
- On the efficiency of algorithms of analysisBulletin of the American Mathematical Society, 1985
- A Limit Theorem for the Norm of Random MatricesThe Annals of Probability, 1980
- On the statistical distribution of the widths and spacings of nuclear resonance levelsMathematical Proceedings of the Cambridge Philosophical Society, 1951
- Numerical inverting of matrices of high order. IIProceedings of the American Mathematical Society, 1951