Improved bounds for the nyström method with application to kernel classification

Mathematics

• Power Law 100%
• Matrix (Mathematics) 100%
• Eigenvalue 100%
• Random Matrix Theory 100%
• Data Point 100%
• Approximation Error 100%
• Approximates 50%
• Small Set 50%
• Total Number 50%
• Integral Operator 50%
• Positive Constant 50%
• Support Vector Machine 50%
• Error Bound 50%

Keyphrases

• Kernel Classification 100%
• Power Law 66%
• Eigenvalues 66%
• Random Matrix Theory 66%
• One-based 66%
• Generalization Performance 66%
• Kernel Matrix 66%
• Number of Data 33%
• Small Sets 33%
• Integral Operator 33%
• Classification Approach 33%
• Operator Basis 33%
• Incoherence 33%
• Positive Definite Matrix 33%
• Approximation Error 33%
• Support Vector 33%
• Spectral Norm 33%

Computer Science

• Approximation (Algorithm) 100%
• Eigenvalue 100%
• Generalization Performance 100%
• classification approach 50%
• Positive Constant 50%
• Support Vector 50%