Metric learning: cross-entropy vs. pairwise losses.
Malik BoudiafJérôme RonyImtiaz Masud ZikoEric GrangerMarco PedersoliPablo PiantanidaIsmail Ben AyedPublished in: CoRR (2020)
Keyphrases
- metric learning
- cross entropy
- pairwise
- semi supervised
- maximum likelihood
- log likelihood
- multi class
- language modeling
- loss function
- markov random field
- evaluation metrics
- similarity measure
- error function
- pairwise constraints
- distance metric
- dimensionality reduction
- feature space
- information retrieval
- neural network
- learning tasks