Sparse PLS hyper-parameters optimisation for investigating brain-behaviour relationships.
Fabio S. FerreiraMaria J. RosaMichael MoutoussisRay DolanJohn Shawe-TaylorJohn AshburnerJanaina Mourão MirandaPublished in: PRNI (2018)
Keyphrases
- hyperparameters
- model selection
- cross validation
- closed form
- prior information
- bayesian framework
- random sampling
- support vector
- gaussian process
- bayesian inference
- sample size
- maximum a posteriori
- partial least squares
- variational bayes
- maximum likelihood
- em algorithm
- incremental learning
- posterior distribution
- noise level
- ls svm
- incomplete data
- latent variables
- bayesian methods
- genetic algorithm
- high dimensional
- training set
- parameter space
- active learning
- parameter settings
- data mining
- prior knowledge
- multiscale
- conjugate priors