Monte Carlo Method
Wang-Landau sampling for estimation of the reliability of physical networks
Wanyok Atisattapong, Pasin Marupanthorn
We proposed alternative ways of using the Wang-Landau Sampling to estimate network reliability. The advantages and disadvantages of all schemes were compared.
This project was funded by Thammasat University
Tag: Monte Carlo Method, Wang-Landau Sampling, Network Reliability
Status: Completed, 2019
A 1/t algorithm with the density of two states for estimating
multidimensional integrals
Wanyok Atisattapong, Pasin Marupanthorn
Approximation of an ill-behaved integral is a difficult task even using the Wang-Landau Sampling in the previous work.
ฯwe were successful to speed up and reduce algorithm complexity to estimate both general multidimensional and ill-behaved integrals.
The idea can be readily generalized to other continuous models.
Tag: Monte Carlo Method, Wang-Landau Sampling, Numerical Integration
This project was funded by Thammasat University and won the excellent reseach award in 2019
Status: Completed, 2017
Tag: Monte Carlo Method, Wang-Landau Sampling, Numerical Integration
Obviating the bin width effect of the 1/t algorithm formultidimensional numerical integration
The Wang-Landau Sampling is very efficient to calculate the density of states in statistical physics systems. It was applied to calculate a multidimensional integral. However, the algorithm faced with error saturation. In this work, we found that the error saturation results from fixing bin width during grid discretizaion. We were successful to eliminate the error saturation in the Wang-Landau sampling by obviating the bin width effect.
This project was funded by Thammasat University and Center of Excellent in Mathematics (CEM)
Status: Completed, 2016
Wanyok Atisattapong, Pasin Marupanthorn