The Bias-Variance Dilemma of the Monte Carlo Method
We investigate the setting in which Monte Carlo methods are used and draw a parallel to the formal setting of statistical inference. In particular, we find that Monte Carlo approximation gives rise to a bias-variance dilemma. We show that it is possible to construct a biased approximation scheme with a lower approximation error than a related unbiased algorithm.
Unable to display preview. Download preview PDF.
- 2.W. R. Gilks, S. Richardson, and D. J. Spiegelhalter, editors. Markov Chain Monte Carlo in Practice. Chapman&Hall, 1996.Google Scholar
- 3.W. Hardle. Applied nonparametric regression. Cambridge University Press, 1990.Google Scholar
- 5.R. M. Neal. Bayesian Learning for Neural Networks. Springer-Verlag, 1996.Google Scholar
- 6.R. M. Neal. Annealed importance sampling. Technical Report No. 9805, Dept. of Statistics, University of Toronto, 1998.Google Scholar