Abstract
Bayesian Networks structure learning (BNSL) is a troublesome problem that aims to search for an optimal structure. An exact search tends to sacrifice a significant amount of time and memory to promote accuracy, while the local search can tackle complex networks with thousands of variables but commonly gets stuck in a local optimum. In this paper, two novel and practical operators and a derived operator are proposed to perturb structures and maintain the acyclicity. Then, we design a framework, incorporating an influential perturbation factor integrated by three proposed operators, to escape current local optimal and improve the dilemma that outcomes trap in local optimal. The experimental results illustrate that our algorithm can output competitive results compared with the state-of-the-art constraint-based method in most cases. Meanwhile, our algorithm reaches an equivalent or better solution found by the state-of-the-art exact search and hybrid methods.
Highlights
A crucial matter in artificial intelligence is the development of models learned from data that can provide a structural representation based on the domain knowledge and have their own definite semantics
The Bayesian networks (BNs) learning problem consists of the directed acyclic graph (DAG) recovery and numerical parameters characterizing it, called BN structure learning (BNSL) and BN parameter learning (BNPL), respectively
A new framework, ILSM, based on Iterated Local Search (ILS), is proposed that incorporates the Momentum factor integrated by three proposed operators
Summary
A crucial matter in artificial intelligence is the development of models learned from data that can provide a structural representation based on the domain knowledge and have their own definite semantics. BN uses a graphed-based representation as the basis for compactly encoding a complex distribution over a high-dimensional space It is composed of a directed acyclic graph (DAG) where nodes correspond to the variables, and the directed edges correspond to direct probabilistic interactions between them. The BN learning problem consists of the DAG recovery and numerical parameters characterizing it, called BN structure learning (BNSL) and BN parameter learning (BNPL), respectively The former is considered more challenging than the latter because searching for an accurate graphical representation is NP-hard [7]. There are hybrid approaches, which first prune the search space using a constraint-based method, followed by a greedy search for an optimal structure. The second one, called local search, is generally based on heuristics over the DAG space, such as hill-climbing (HC) [22]. BN can be considered as a recipe for factorizing a joint distribution of {X1, X2, . . . , Xn}: n
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
