Combining a relaxed EM algorithm with Occam's razor for Bayesian variable selection in high-dimensional regression.
Pierre LatouchePierre-Alexandre MatteiCharles BouveyronJulien ChiquetPublished in: J. Multivar. Anal. (2016)
Keyphrases
- variable selection
- em algorithm
- high dimensional
- bayesian model selection
- maximum likelihood
- model selection
- expectation maximization
- linear models
- mixture model
- parameter estimation
- cross validation
- hyperparameters
- gaussian processes
- posterior distribution
- gaussian mixture model
- dimension reduction
- high dimensional data
- maximum a posteriori
- generative model
- maximum likelihood estimation
- regression model
- data points
- posterior probability
- relevance vector machine
- high dimensionality
- probability density function
- low dimensional
- nearest neighbor
- probabilistic model
- feature space
- machine learning
- object recognition