Supercomputers, Monte Carlo Simulation and Regression Analysis

摘要

The paper focuses on the use of supercomputers in Monte Carlo experiments with multivariate regression analysis. Supercomputers provide a new tool for management scientists. The application of this tool requires thinking in parallel or vector mode. The parallel mode needs to exploit a specific dimension of the Monte Carlo experiment (namely the replicates of that experiment). Then Ordinary Least Squares on a CYBER 205 takes only 1.4% of the time needed on a VAX 8700. Estimated Generalized Least Squares, however, is slower on the CYBER 205 because it requires matrix inversion.