Learning Bayesian network parameters from small data sets: A further constrained qualitatively maximum a posteriori method.
Zhi-gao GuoXiao-Guang GaoHao RenYu YangRuo-hai DiDa-Qing ChenPublished in: Int. J. Approx. Reason. (2017)
Keyphrases
- maximum a posteriori
- parameter learning
- em algorithm
- maximum likelihood
- small data sets
- hyperparameters
- parameter estimation
- prior distribution
- expectation maximization
- cost function
- markov random field
- structural learning
- energy function
- computational complexity
- map estimation
- generalized gaussian
- bayesian framework
- bayesian networks
- prior model
- reinforcement learning
- prior knowledge
- image reconstruction
- data sets
- probabilistic model
- pairwise
- prior information
- image registration