Accession Number : ADA292874

Title :   Theory & Numerical Methods for Problems in Shape Reconstruction and Large Deviations.

Descriptive Note : Final technical rept. 1 Jun 93-31 May 94,

Corporate Author : BROWN UNIV PROVIDENCE RI DIV OF APPLIED MATHEMATICS

Personal Author(s) : Dupuis, Paul

PDF Url : ADA292874

Report Date : 26 JUL 1994

Pagination or Media Count : 5

Abstract : This report describes work on problems from stochastic systems theory where the noise effects are small. The work has addressed two current needs: extension of the theory to cover practically interesting problems (such as design of data networks), and the development of computational methods for large deviation problems. Problems of particular interest that were dealt with are approximation of rare events in queueing systems and the control of reliable systems. We also discuss a recently developed method for obtaining rate of convergence estimates for a variety of approximations in deterministic and stochastic optimal control. (AN)

Descriptors :   *QUEUEING THEORY, *STOCHASTIC CONTROL, MATHEMATICAL MODELS, ALGORITHMS, DATA PROCESSING, OPTIMIZATION, FINITE DIFFERENCE THEORY, RELIABILITY, APPROXIMATION(MATHEMATICS), COMMUNICATIONS NETWORKS, CONVERGENCE, NOISE, CONTROL THEORY, COMPUTER NETWORKS, NUMERICAL METHODS AND PROCEDURES, MARKOV PROCESSES.

Subject Categories : Numerical Mathematics
      Operations Research

Distribution Statement : APPROVED FOR PUBLIC RELEASE