An Asymptotically Optimal Approximation of the Conditional Mean Channel Estimator based on Gaussian Mixture Models.
Michael KollerBenedikt FeslNurettin TuranWolfgang UtschickPublished in: CoRR (2021)
Keyphrases
- asymptotically optimal
- gaussian mixture model
- maximum likelihood
- asymptotic optimality
- mixture model
- em algorithm
- speaker recognition
- heavy traffic
- feature vectors
- expectation maximization
- speaker identification
- arrival rate
- maximum a posteriori
- density estimation
- probability density function
- optimal policy
- feature space
- information systems
- hyperparameters
- queueing networks
- probability distribution
- variational bayes
- feature extraction