Search papers, labs, and topics across Lattice.
The authors tackle structural consensus fusion of multiple Bayesian networks under strict treewidth constraints to guarantee computationally tractable inference. Unrestricted model aggregation typically results in dense directed graphs where exact probabilistic inference is NP-hard, necessitating bounded-complexity structural regularization. Experimental results show that the proposed genetic search efficiently preserves core structural dependencies from input ensembles while strictly maintaining the predefined treewidth bound.
Multi-source Bayesian network fusion usually destroys inference tractability, but imposing hard treewidth constraints via evolutionary search preserves consensus graph structure without sacrificing computational feasibility.
This paper introduces an evolutionary computation approach for consensus in structural Bayesian Network (BN) fusion under the constraint of limited treewidth. The consensus BN aims to reconcile multiple input BNs into a single one that retains key structural features present in the original networks. Treewidth, a graph-based parameter associated with computationally tractable inference, is utilized to restrict the complexity of the resulting network. A genetic algorithm is proposed to look for a BN that codifies as much information about the unrestricted fusion as possible while ensuring the treewidth restriction. Experimental evaluation demonstrates the genetic algorithm's ability to obtain consensus BNs with limited treewidth, providing a valuable tool for aggregating information from diverse sources while returning a computationally actionable model.
