Physics – Condensed Matter – Disordered Systems and Neural Networks
Scientific paper
2011-12-20
Physics
Condensed Matter
Disordered Systems and Neural Networks
18 pages, 6 figures
Scientific paper
I consider mean-field approximations for solving two fundamental problems in Ising models, namely estimating 2-point correlations given the couplings (direct problem) and estimating the couplings given the correlations and the magnetizations (inverse problem). I derive analytical expressions for the Bethe approximation, which allow to solve the problems without running the Susceptibility Propagation algorithm. I compare the accuracy of 5 different mean field approximations on several models (diluted ferromagnets and spin glasses) on random graphs and regular lattices. A simple improvement over these approximations is proposed.
No associations
LandOfFree
On mean-field approximations for estimating correlations and solving the inverse Ising problem does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with On mean-field approximations for estimating correlations and solving the inverse Ising problem, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and On mean-field approximations for estimating correlations and solving the inverse Ising problem will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-188684