On loss functions which minimize to conditional expected values and posterior proba- bilities.
John W. MillerRodney M. GoodmanPadhraic SmythPublished in: IEEE Trans. Inf. Theory (1993)
Keyphrases
- loss function
- expected values
- pairwise
- support vector
- loss minimization
- learning to rank
- squared error
- risk minimization
- reproducing kernel hilbert space
- probability distribution
- bayesian framework
- posterior distribution
- convex loss functions
- generative model
- model selection
- marginal distributions
- bregman divergences
- probabilistic model
- active learning