<t>> "I assume I need to be changing the direction numbers"I wouldn't play with changing or generation of the direction numbers. The direction numbers are the key to the efficiency of the Sobol' sequence generator. You can readI. Sobol', D. Asotsky, A. Kreinin, S. Kucherenko. Construction and Compar...

QuoteOriginally posted by: TinManBear in mind too that you can't use Cholesky with SobolWhy is that ?

For statistical and other test of BRODA's and other Sobol' sequence generators you can read I. Sobol', D. Asotsky, A. Kreinin, S. Kucherenko. Construction and Comparison of High-Dimensional Sobol' Generators, 2011, Wilmott Journal, Nov, pp. 64-79

<t>..and two other papers on application of Global sensitivity analysis in finance:1. M. Bianchetti, S. Kucherenko, S. Scoleri, Better Pricing and Risk Management with High Dimensional Quasi Monte Carlo, WBS 10th Fixed Income Conference, Barcelona , September, 2014;2. Kucherenko S., Shah N. The Impo...

<t>A note of caution: the easiness of implementation doesn't imply efficiency: good Sobol' sequences are very difficult to construct. For comparison of various Sobol' sequence generators used in the industry read I. Sobol', D. Asotsky, A. Kreinin, S. Kucherenko. Construction and Comparison of High-D...

<t>Quote Would be interesting to pin down exactly the cases where the brownian bridge discretisation works less well than the standard one (you claim cliquets?) as basically I've always found sobol unusable without bb. Also, have you tried the spectral construction of the wiener process? Should prob...

<t>Quotehave done any tests for the LMM? eg a TARN cf the examples in More Mathematical Finance.No, we haven't done such tests but we would like to do them. Would you be interested in doing them together ?With colleagues we've done recently a work ( Marco Bianchetti, Sergei Kucherenko, Stefano Scole...

Difficult to tell why there is a saturation for many SobolSeq generators at low dimensions ( Dim = 8 ) at Fig. 1. I would be interested to see other similar comparisons with other discrepancy measures ( not just 0.0 anchored L2 discrepancy ).

<t>Currently, the Sobol' sequence generator with the highest dimension is BRODA's SobolSeq32000. It satisfy additional uniformity properties: Property A for all 32000 dimensions and Property A' for 5 adjacent dimensions. For comparison of various Sobol' sequence generators used in the industry read ...

<t>Quotei haven't done extensive tests but I reckon that JoeKuo D7 are the best. For comparison of various Sobol' sequence generators used in the industry ( including Joe&Kuo ) read I. Sobol', D. Asotsky, A. Kreinin, S. Kucherenko. Construction and Comparison of High-Dimensional Sobol' Generator...

<r>QuoteOriginally posted by: animeshsaxenaQuoteI am using the direction numbers given by this website <URL url="http://web.maths.unsw.edu.au/~fkuo/sobol/index.html">http://web.maths.unsw.edu.au/~fkuo/sobol/index.html</URL> I understand it's widely used so direction numbers are bound to be correct!F...

<t>QuoteOriginally posted by: quartzQuoteBesides that the sobol' direction numbers that have been used for a few years had some flows...Indeed, efficient direction numbers is the key to the perfomance of Sobol' sequence generators. For a comparison of various Sobol' sequence generators used in the i...

<t>QuoteWell, if I look at the results in the book of Jackel, it seems that beyond dimension 15 the difference between pseudo-random (Mersenne Twister) and Sobol numbers is negligible. So if you are building a model for multiple assets, the question is whether this is worth the trouble... Yes, it is...

QuoteI particularly am very interested in global optimization problem.Sobol' sequences can be efficiently used for solving global optimization problems. Read thesepapers for details

<t>Mersenne twister is just another random number generator. It gives a Monte Carlo convergence rate 1/sqrt(N), N is the number of sampled points or paths ).Sobol' sequence is the best ( in practice ) low-discrepancy sequence. It provides a rate of convergence 1/N ( for path dependent options it is ...