Numerical Techniques for Stochastic Systems: A Collection of Papers Based on the Lectures Presented at the Conference on Numerical Techniques for Stochastic Systems, Held at Gargnano, Italy, September 1979

This conference proceedings contains 23 papers on the estimation, identification and modeling of stochastic systems and the stochastic design and analysis of numerical algorithms. Topics discussed include nonlinear stochastic system identification as an optimization problem, fast algorithms for recursive estimation and identification, the application of spectral decomposition to system identification, a continuous-time approximation to optimal nonlinear filtering, the use of stochastic models in air pollution prediction and control, stochastic processes in the modeling of earthquake occurrence, maximum likelihood estimates of errors ARMA regression models, stochastic analysis of in numerical processes, the advantages of parallel computation in numerical optimization, a probabilistic framework for combinatorial optimization, and a simple Bayesian algorithm for multidimensional global optimization.