A Fixed-Point Algorithm for Finding the Optimal Covariance Matrix in Kernel Density Modeling.
José M. Leiva-MurilloAntonio Artés-RodríguezPublished in: ICASSP (5) (2006)
Keyphrases
- fixed point
- covariance matrix
- dynamic programming
- optimal solution
- kernel density
- objective function
- np hard
- worst case
- simulated annealing
- probabilistic model
- computational complexity
- similarity measure
- search space
- model free
- reinforcement learning
- hidden markov models
- principal component analysis
- em algorithm
- sample size
- feature space
- convergence rate
- policy iteration
- geometrical interpretation