Variable selection via the composite likelihood method for multilevel longitudinal data with missing responses and covariates.
Haocheng LiDi ShuWenqing HeGrace Y. YiPublished in: Comput. Stat. Data Anal. (2019)
Keyphrases
- membership functions
- input variables
- fuzzy logic
- variable selection
- cross validation
- pairwise
- dimension reduction
- linear models
- neural network
- objective function
- high dimensional
- model selection
- canonical correlation analysis
- ls svm
- missing data
- support vector machine svm
- unsupervised learning
- maximum likelihood
- probabilistic model
- preprocessing
- object recognition
- pattern recognition
- feature extraction