Maximum Mean Discrepancy for Class Ratio Estimation: Convergence Bounds and Kernel Selection.
Arun Shankar IyerJ. Saketha NathSunita SarawagiPublished in: ICML (2014)
Keyphrases
- parzen window
- finite sample
- convergence rate
- pre selected
- rates of convergence
- upper bound
- kernel function
- support vector
- class separability
- kernel density estimation
- reproducing kernel hilbert space
- density estimation
- lower bound
- convergence speed
- finite number
- learning machines
- kernel matrix
- multiple kernel learning
- upper and lower bounds
- selection criterion
- class labels
- linear svm
- parameter estimation
- markov random field
- feature space
- initial guess
- feature selection