A Mutually-Dependent Hadamard Kernel for Modelling Latent Variable Couplings.
Sami RemesMarkus HeinonenSamuel KaskiPublished in: ACML (2017)
Keyphrases
- latent variables
- mutually dependent
- random variables
- probabilistic model
- latent variable models
- kernel function
- latent space
- hierarchical model
- gaussian process
- topic models
- real valued
- graphical models
- prior knowledge
- hidden variables
- feature space
- kernel methods
- probability distribution
- support vector
- latent structure
- hadamard transform
- learning algorithm
- probabilistic latent semantic analysis
- pairwise
- bayesian networks
- structured prediction
- observed variables