Gaussian process hyper-parameter estimation using Parallel Asymptotically Independent Markov Sampling.
Alfredo Garbuno-InigoFrancisco Alejandro DiazDelaOKonstantin ZuevPublished in: Comput. Stat. Data Anal. (2016)
Keyphrases
- parameter estimation
- gaussian process
- model selection
- sample size
- hyperparameters
- gaussian processes
- approximate inference
- markov chain monte carlo
- regression model
- cross validation
- random sampling
- gaussian process regression
- maximum likelihood
- random fields
- statistical models
- likelihood function
- markov fields
- bayesian framework
- parameter estimation algorithm
- machine learning
- least squares
- latent variables
- markov random field
- em algorithm
- semi supervised
- feature selection
- parameter values
- posterior distribution
- worst case
- markov chain
- expectation maximization
- higher order
- structure learning
- upper bound
- optical flow
- pairwise
- training data