Systems and methods for causal inference in network structures using belief propagation
US11068799B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Sep 3, 2015 |
| Grant date | Jul 20, 2021 |
| Priority date | — |
| Expiry date | Jun 16, 2038 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N7/00
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Systems and method for perturbing a system include obtaining directed acyclic/cyclic graph candidates {GI, . . . , GN} for the system. Each Gi in {Gj, . . . GN} includes a causal relationship between a parent and child node. {GI, GN} demonstrate Markov equivalence. Observed data D is obtained for the nodes. For each respective Gi, the marginal probability of a parent node xi in Gi is clamped by D while computing a distribution of marginal probabilities for a child node yi, by Bayesian network or Dynamic Bayesian network belief propagation using an interaction function. The observed distribution for the child node yi, in D and the computed distribution of marginal probabilities for the child node yi are scored using a nonparametric function, and such scores inform the selection of a directed/cyclic graph from {GI, . . . , GN}. The system is perturbed using a perturbation that relies upon a causal relationship in the selected directed acyclic/cyclic graph.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.