A two phase approach to Bayesian network model selection and comparison between the MDL and DGM scoring heuristics
نویسندگان
چکیده
This paper presents an eficient algorithm for learning a Bayesian belief network (BBN) structure from a database, as well as providing a comparison between two BBN structure fitness functions. A Bayesian belief network is a directed acyclic graph representing conditional expectations. In this paper, we propose a two-phase algorithm. The first phase uses asymptotically correct structure learning for eficient search space exploration, while the second phase uses greedy model selection for accurate search space exploitation. The minimum description length (MDL) structure fitness function is also compared with the database given model probability (DGM) fitness function in the second phase. The model selection algorithms are applied to the ALARM network to provide a comparison for the accuracy of the techniques.
منابع مشابه
Project Portfolio Risk Response Selection Using Bayesian Belief Networks
Risk identification, impact assessment, and response planning constitute three building blocks of project risk management. Correspondingly, three types of interactions could be envisioned between risks, between impacts of several risks on a portfolio component, and between several responses. While the interdependency of risks is a well-recognized issue, the other two types of interactions remai...
متن کاملOn the importance of using treewidth as a model-selection criterion for learning Bayesian networks
This paper is motivated by the desire to learn Bayesian networks that allow efficient inference. Traditionally, model selection criteria such as BIC/MDL focus on learning Bayesian networks that fit the data and have low representation complexity (i.e. the number of parameters needed to specify the network). However, these criteria do not take into account the complexity of inference in the resu...
متن کاملDesign of An Integrated Robust Optimization Model for Closed-Loop Supply Chain and supplier and remanufacturing subcontractor selection
The development of optimization and mathematical models for closed loop supply chain (CLSC) design has attracted considerable interest over the past decades. However, the uncertainties that are inherent in the network design are challenging the capabilities of the developed tools. In CLSC Uncertainty in demand is major source of uncertainty. The aim of this paper, therefore, is to present a Rob...
متن کاملA Novel Hybrid Fuzzy Multi-Criteria Decision-Making Model for Supplier Selection Problem (A Case Study in Advertising industry)
Choosing the proper supplier has a critical role in design of supply chain. This problem is complicated because each supplier may fulfills some of the manufacturer criteria and choosing the best supplier is a Multiple-Criteria Decision Making problem. This paper proposes a novel hybrid approach to rank suppliers in advertising industry and considers two new criteria to evaluate the suppliers in...
متن کاملA Particle Swarm Optimization and Immune Theory-Based Algorithm for Structure Learning of Bayesian Networks
Bayesian network is a directed acyclic graph. Existing Bayesian network learning approaches based on search & scoring usually work with a heuristic search for finding the highest scoring structure. This paper describes a new data mining algorithm to learn Bayesian networks structures based on an immune binary particle swarm optimization (IBPSO) method and the Minimum Description Length (MDL) pr...
متن کامل