An algorithm inspired by the deterministic annealing approach to avoid local minima in artificial potential fields.
Nara Strappa Facchinetti DoriaEduardo Oliveira FreireJoão Carlos BasilioPublished in: ICAR (2013)
Keyphrases
- learning algorithm
- expectation maximization
- deterministic annealing
- simulated annealing
- objective function
- k means
- cost function
- dynamic programming
- probabilistic model
- data clustering
- similarity measure
- neural network
- bayesian framework
- data mining
- worst case
- em algorithm
- cluster analysis
- image segmentation
- globally optimal