Improved LDA Classifier Based on Spiked Models

Mathematics

• Discriminant Analysis 100%
• Covariance Matrix 80%
• Gaussian Mixture Model 40%
• Regularization 20%
• Synthetic Data 20%
• Inverse Covariance Matrix 20%
• Training Data 20%
• Identitymatrix 20%
• Numerical Experiment 20%
• Real Data 20%

Keyphrases

• Robust Linear Discriminant Analysis 100%
• Spiked Model 100%
• Regularized LDA 50%
• LDA Classifier 25%
• Low-rank Perturbation 25%