Investigating maximum likelihood based training of infinite mixtures for uncertainty quantification.
Sina DäubenerAsja FischerPublished in: CoRR (2020)
Keyphrases
- maximum likelihood
- mixture model
- expectation maximization
- em algorithm
- discriminative training
- training set
- training samples
- online learning
- maximum a posteriori
- incomplete information
- uncertain data
- inherent uncertainty
- training phase
- parameter estimation
- training process
- gaussian distribution
- source separation
- supervised learning