Output-weighted and relative entropy loss functions for deep learning precursors of extreme events.
Samuel H. RudyThemistoklis P. SapsisPublished in: CoRR (2021)
Keyphrases
- deep learning
- loss function
- relative entropy
- bregman divergences
- unsupervised learning
- information theoretic
- pairwise
- machine learning
- log likelihood
- cost sensitive
- support vector
- maximum entropy
- mahalanobis distance
- mutual information
- covariance matrix
- boosting algorithms
- exponential family
- nearest neighbor
- high dimensional
- order statistics
- kl divergence
- theoretical guarantees
- multi class