A finite-step global convergence algorithm for the cumulant-based parameter estimation of multichannel moving average processes.
Lang TongYujiro InouyeRuey-Wen LiuPublished in: ICASSP (1991)
Keyphrases
- parameter estimation
- expectation maximization
- global convergence
- em algorithm
- convergence rate
- least squares
- parameter estimation algorithm
- optimization method
- global optimum
- convergence speed
- moving average
- model selection
- markov random field
- optimization methods
- maximum likelihood
- maximum likelihood estimation
- k means
- convergence analysis
- probabilistic model
- optimal solution
- objective function
- convex minimization
- constrained optimization
- neural network
- ant colony optimization
- optimization algorithm
- hidden markov models
- cost function
- search algorithm
- similarity measure
- genetic algorithm
- data mining