It is all in the noise: Efficient multi-task Gaussian process inference with structured residuals.
Barbara RakitschChristoph LippertKarsten M. BorgwardtOliver SteglePublished in: NIPS (2013)
Keyphrases
- gaussian process
- gaussian processes
- multi task
- multi task learning
- regression model
- learning tasks
- semi supervised
- model selection
- latent variables
- approximate inference
- bayesian framework
- noise reduction
- noise level
- bayesian inference
- exact inference
- data sets
- hyperparameters
- learning problems
- em algorithm
- high dimensional
- feature selection