Probabilistic Inference in Credal Networks: New Complexity Results

被引:25
作者
Maua, Denis Deratani [1 ]
de Campos, Cassio P. [2 ]
Benavoli, Alessio [2 ]
Antonucci, Alessandro [2 ]
机构
[1] Univ Sao Paulo, Escola Politecn, BR-05508010 Sao Paulo, Brazil
[2] Ist Dalle Molle Studi Intelligenza Artificiale, CH-6928 Manno, Switzerland
基金
巴西圣保罗研究基金会; 瑞士国家科学基金会;
关键词
SETS;
D O I
10.1613/jair.4355
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Credal networks are graph-based statistical models whose parameters take values in a set, instead of being sharply specified as in traditional statistical models (e.g., Bayesian networks). The computational complexity of inferences on such models depends on the irrelevance/independence concept adopted. In this paper, we study inferential complexity under the concepts of epistemic irrelevance and strong independence. We show that inferences under strong independence are NP-hard even in trees with binary variables except for a single ternary one. We prove that under epistemic irrelevance the polynomial-time complexity of inferences in credal trees is not likely to extend to more general models (e.g., singly connected topologies). These results clearly distinguish networks that admit efficient inferences and those where inferences are most likely hard, and settle several open questions regarding their computational complexity. We show that these results remain valid even if we disallow the use of zero probabilities. We also show that the computation of bounds on the probability of the future state in a hidden Markov model is the same whether we assume epistemic irrelevance or strong independence, and we prove a similar result for inference in naive Bayes structures. These inferential equivalences are important for practitioners, as hidden Markov models and naive Bayes structures are used in real applications of imprecise probability.
引用
收藏
页码:603 / 637
页数:35
相关论文
共 55 条
[1]  
[Anonymous], 2010, P BRIT MACH VIS C
[2]  
[Anonymous], 1994, P 5 INT C INF PROC M
[3]  
[Anonymous], P 19 EUR C ART INT E
[4]  
[Anonymous], 1974, Stochastic Geometry
[5]  
[Anonymous], STAT ENG INFORM SCI
[6]  
[Anonymous], P 8 INT S IMPR PROB
[7]  
[Anonymous], DMV SEM
[8]  
[Anonymous], 2004, P 20 C UNC ART INT
[9]  
[Anonymous], 1979, Computers and Intractablity: A Guide to the Theory of NP-Completeness
[10]  
[Anonymous], 2013, P 16 C INF FUS FUSIO