A convex parametrization of a new class of universal kernel functions

Mathematics

• Universal Function 100%
• Kernel Function 79%
• Parametrization 73%
• kernel 50%
• Class 22%
• Learning Algorithm 11%
• Parameterization 9%
• Scalability 6%
• Support Vector Machine 5%
• Positive Matrices 5%
• Tractability 5%
• Margin 5%
• Training 4%
• Optimise 4%
• Bandwidth 4%
• Neural Networks 4%
• Learning 4%
• Testing 3%
• Imply 2%
• Polynomial 2%

Engineering & Materials Science

• Parameterization 52%
• Set theory 47%
• Learning algorithms 41%
• Scalability 20%
• Support vector machines 19%
• Polynomials 19%
• Bandwidth 15%
• Neural networks 15%
• Testing 12%