Bootstrap estimation and model selection for multivariate normal mixtures using parallel computing with graphics processing units.
Masanari IidaYoichi MiyataTakayuki ShiohamaPublished in: Commun. Stat. Simul. Comput. (2018)
Keyphrases
- model selection
- parallel computing
- graphics processing units
- cross validation
- multivariate normal
- parameter estimation
- mixture model
- parallel programming
- massively parallel
- processing units
- selection criterion
- parallel computation
- high performance computing
- hypothesis tests
- sample size
- computing systems
- hyperparameters
- shared memory
- commodity hardware
- machine learning
- gpu implementation
- covariance matrix
- density estimation
- general purpose
- generalization error
- gaussian mixture
- gaussian mixture model
- compute unified device architecture
- covariance matrices
- feature selection
- higher order
- parallel machines
- floating point
- real time
- parallel processing
- parallel algorithm
- expectation maximization
- probabilistic model
- similarity measure
- data mining