A Unified Perspective on Natural Gradient Variational Inference with Gaussian Mixture Models.
Oleg ArenzPhilipp DahlingerZihan YeMichael VolppGerhard NeumannPublished in: CoRR (2022)
Keyphrases
- variational inference
- gaussian mixture model
- mixture model
- density estimation
- em algorithm
- bayesian inference
- probabilistic model
- expectation maximization
- exponential family
- posterior distribution
- maximum likelihood
- generative model
- model selection
- unsupervised learning
- latent dirichlet allocation
- language model
- topic models
- gaussian distribution
- gaussian process
- closed form
- multiscale
- hyperparameters
- independent component analysis
- parameter estimation
- probability density function
- maximum a posteriori
- data mining
- feature vectors
- pattern recognition
- bayesian networks
- computer vision
- learning algorithm