Branch and Bound with Mini-Bucket Heuristics
نویسندگان
چکیده
The paper describes a new generic branch and bound scheme that uses heuristics generated mechanically by the mini-bucket approximation. The scheme is presented and evaluated for the Most Probable Explanation (MPE) task in Bayesian networks. We show that the mini-bucket scheme yields monotonic heuristics of varying strengths which cause different amounts of pruning during search. The resulting Branch and Bound with Mini-Bucket algorithm (BBMB), is evaluated using random networks as well as coding and medical diagnosis networks. The results show that the BBMB scheme overcomes the memory explosion of bucket-elimination and is applicable for a large range of problems because of its gradual tradeoff of space for time and of time for accuracy.
منابع مشابه
Mini-Bucket Heuristics for Improved Search
The paper is a second in a series of two pa pers evaluating the power of a new scheme that generates search heuristics mechanically. The heuristics are extracted from an approx imation scheme called mini-bucket elimina tion that was recently introduced. The first paper introduced the idea and evaluated it within Branch-and-Bound search. In the cur rent paper the idea is further extended and...
متن کاملEmpowering Mini-Bucket in Anytime Heuristic Search with Look-Ahead: Preliminary Evaluation
The paper explores the potential of look-ahead methods within the context of AND/OR search in graphical models using the Mini-Bucket heuristic for combinatorial optimization tasks (e.g., weighted CSPS or MAP inference). We study how these methods can be used to compensate for the approximation error of the initially generated Mini-Bucket heuristics, within the context of anytime Branch-And-Boun...
متن کاملBeyond Static Mini-Bucket: Towards Integrating with Iterative Cost-Shifting Based Dynamic Heuristics
We explore the use of iterative cost-shifting as a dynamic heuristic generator for solving MPE in graphical models via Branch and Bound. When mini-bucket elimination is limited by its memory budget, it may not provide good heuristics. This can happen often when the graphical model has a very high induced width with large variable domain sizes. In addition, we explore a hybrid setup where both M...
متن کاملUsing Mini-Bucket Heuristics for Max-CSP
This paper evaluates the power of a new scheme that generates search heuristics mechanically. This approach was presented and evaluated rst in the context of optimization in belief networks. In this paper we extend this work to Max-CSP. The approach involves extracting heuristics from a parameterized approximation scheme called Mini-Bucket elimination that allows controlled trade-o between comp...
متن کاملAND/OR Branch-and-Bound for Probabilistic Inference
The Most Probable Explanation problem asks for the most probable assignment on random variables of a Bayesian network, when the values of some variables are known. We present the AND/OR search technique for solving the problem, combined with heuristic Branch-and-Bound pruning. We also present the mini-bucket elimination technique for bounded inference and show how it can be used to produce heur...
متن کامل