Parallelized formulation of the maximum likelihood-expectation maximization algorithm for fine-grain message-passing architectures.
José Cruz-RiveraEdward V. R. Di BellaD. Scott WillsThomas K. GaylordElias N. GlytsisPublished in: IEEE Trans. Medical Imaging (1995)
Keyphrases
- expectation maximization
- maximum likelihood
- em algorithm
- message passing
- bayesian framework
- parameter estimation
- fine grain
- maximum a posteriori
- dynamic programming
- optimal solution
- loopy belief propagation
- probabilistic model
- np hard
- k means
- computational complexity
- similarity measure
- preprocessing
- parallel computation
- sum product algorithm
- linear programming
- highly efficient
- junction tree
- inference in graphical models