Quasi-Random Sequences and Their Discrepancies
Download
Report
Transcript Quasi-Random Sequences and Their Discrepancies
Quasi-Random Sequences and Their
Discrepancies
Author: William J. Morokoff & Russel E. Caflisch
Presented by Shuaiyuan Zhou
Oct-16-2009
Monte Carlo Integration
Quasi-Random Sequences and their Discrepancy
Koksma-Hlawka inequality
Wozmiakowski identity
Low discrepancy sequences
L2 discrepancy
Orthogonal projections
Computational speed
Monte Carlo integration
In mathematics, Monte Carlo integration is numerical
integration using random numbers.
When we evaluate the integral of a function, f(x), in the s-
dimensional unit cube, Is, we are in fact calculating the
average of the function at a set of randomly sampled points.
The accuracy of Monte Carlo integration lies in the selection
of the sequence of points {xn}.
Random sequences
Quasi-random sequences (low discrepancy sequences)
Quasi-random sequences
Sequences with good uniformity properties
Uniformity of a sequence is measured by its discrepancy: the
error in representation of the volume of subsets of the unit
cube by the fraction of points in the subsets.
Discrepancy
For a set of J in Is and a sequence of N points {xn} in Is,
define
If E is the set of all subrectangles of Is and E* is the set of all
subrectangles with one corner at 0, then discrepancies and
star discrepancies are defined as
norm:
L2 norm:
Properties of discrepancy
1.
with probability 1, for any dimension s. size of discrepancy.
This bound can be improved !!
2.
This is regarded as the best bound. Quasi-random sequences with this property
are presented below.
3.
E* is a subset of E, while any set in E can be written as a combination of 2s sets
in E*.
4.
nomrs are larger than L2 norms.
Koksma-Hlawka inequality
Use discrepancy as an error bound for Monte Carlo integration.
In one dimension,
In s-dimension, similarly,
Define for all k ≤ s and all sets of k integers 1 ≤ i1 < i2 < … < ik ≤ s, V is the
variation and D* is the star discrepancy of the orthogonal projection of the
sequence {xn} on to the appropriate k-dimensional subspace of Is.
Wozmiakowski identity
Koksma-Hlawka inequality is an over-estimation
L2 star discrepancy and integration error. The average is taken
with respect to the Brownian sheet measure.
Low discrepancy sequences
Discrepancy is of size O((logN)s/N).
Halton
Sobol’
Faure
Halton
the constant grows super-exponentially with dimension
Sobol’ & Faure
For Sobol’ and Faure sequences, bound of discrepancies looks similarly;
with different coefficient of the (logN)sN-1 term.
Sobol’
Since t is a function of s, so the constant grows super-exponentially with
s
Faure
CsF is smaller than both CsS and CsH, and it goes to zero as dimension
increases while the other go to infinity.
Faure sequence is superior.
Is this discrepancy bound good
enough?
1.Discrepancy for a uniform sequence decreases as N increases;
however, the bound has the maximum when N = es.
This means that we can use the bound as a measure of
performance only after its maximum has been attained. Thus, in
high dimensions, a very large number of points is need.
2. Terms other than the leading order one can not be neglected.
3. Even in lower dimension, such as 4 and 16, the discrepancy
bound is not accurate.
L2 discrepancy
Calculation formula for TN and TN *:
TN* > TN for all sequences and all N. The qualitative behavior
of the two descrepancies is similar for large N, but TN is
smoother and has a shorter, less extreme, transient region.
Disadvantage of using TN: no direct connection has been
established between it and integration error.
However, computation of TN indicates that it is a useful gauge
of integration error convergence as a function of N.
Some features about TN:
1. When N is large enough, TN should have a local minimum
2. In high dimensions, unless one uses a very large number of
points, TN is of size N-1.
3. The value of TN is insensitive to where the sequence
begins.
Orthogonal projections
Look at two-dimensional orthogonal projections of the points
in Is.
Choosing different dimension, the result might be quite
different.
How to improve?
Scrambling procedure proposed by Braaten and Weller.
For example, if N points in Is were required, then s sequences
of N random numbers were generated and sorted from
smallest to largest. This mapping of original position in the
sequence to final position was then used to permute the
sequence.
Computational speed
Computation time is approximately proportional to
dimension and to number of points used.
Sobol’ is faster than Halton.
Sobol’ and Halton is faster than Faure.
Thank you !