Belief propagation with multipoint correlations and its application to Inverse problem
Masayuki Ohzeki

TL;DR
This paper introduces explicit formulas for the Bethe approximation incorporating multipoint correlations, leading to a new iterative algorithm that improves accuracy in spin systems and has potential applications in inverse problems.
Contribution
It provides explicit formulas for the Bethe approximation with multipoint correlations and proposes an improved iterative algorithm for better accuracy in spin systems.
Findings
The new algorithm yields more accurate critical point locations.
Formulas include magnetization and correlations in magnetic fields.
Potential application to inverse Ising models.
Abstract
We give explicit formulas of the Bethe approximation with multipoint correlations for systems with magnetic field. The obtained formulas include the closed form of the magnetization and the correlations between adjacent degrees of freedom. On the basis of our results, we propose a new iterative algorithm of the improved Bethe approximation. We confirm that the proposed technique is available for the random spin systems and indeed gives more accurate locations of the critical points. We discuss the possibility of the application of our method to the Inverse Ising model by use of these equations.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
