Visual interpretation of the robustness of Non-Negative Associative Gradient Projection Points over function minimizers in mini-batch sampled loss functions.
Dominic KafkaDaniel N. WilkePublished in: CoRR (2019)
Keyphrases
- loss function
- gradient projection
- reproducing kernel hilbert space
- pairwise
- learning to rank
- squared error
- loss minimization
- logistic regression
- hinge loss
- support vector
- update rules
- decision trees
- risk minimization
- boosting algorithms
- precision and recall
- maximum likelihood
- data points
- training data
- learning algorithm