Linear Shrinkage Estimation of Covariance Matrices Using Low-Complexity Cross-Validation.
Jun TongRui HuJiangtao XiZhitao XiaoQinghua GuoYanguang YuPublished in: CoRR (2018)
Keyphrases
- low complexity
- cross validation
- covariance matrices
- covariance matrix
- model selection
- maximum likelihood
- hyperparameters
- vector space
- computational complexity
- gaussian mixture model
- support vector
- motion estimation
- generalization error
- gaussian distribution
- training set
- distance measure
- linear classifiers
- gaussian mixture
- feature vectors
- closed form
- density estimation
- denoising
- feature selection
- computer vision
- prior knowledge
- knn
- similarity measure
- supervised learning
- image denoising
- objective function
- principal component analysis
- data sets