Expectation-Maximization algorithm for finite mixture of α-stable distributions.
Diego Castillo-BarnesFrancisco Jesús Martínez-MurciaJavier RamírezJ. M. GórrizDiego Salas-GonzalezPublished in: Neurocomputing (2020)
Keyphrases
- expectation maximization
- em algorithm
- learning algorithm
- k means
- times faster
- optimization algorithm
- probabilistic model
- experimental evaluation
- cost function
- detection algorithm
- mixture model
- high accuracy
- dynamic programming
- closed form
- preprocessing
- recognition algorithm
- parameter estimation
- objective function
- gaussian distribution
- tree structure
- segmentation algorithm
- mixture distribution
- linear programming
- particle swarm optimization
- worst case
- np hard
- evolutionary algorithm
- generative model
- simulated annealing
- significant improvement
- matching algorithm
- lower bound
- bayesian framework
- optimal solution
- image segmentation
- gaussian mixture
- computer vision