Batch-Incremental Triplet Sampling for Training Triplet Networks Using Bayesian Updating Theorem.
Milad SikaroudiBenyamin GhojoghFakhri KarrayMark CrowleyHamid R. TizhooshPublished in: CoRR (2020)
Keyphrases
- batch mode
- markov fields
- training set
- bayesian networks
- batch learning
- training process
- social networks
- supervised learning
- data driven
- quotient space
- echo state networks
- incremental update
- unsupervised image segmentation
- heterogeneous networks
- complex networks
- parameter estimation
- markov chain monte carlo
- bayesian inference
- svm training
- recurrent networks
- recurrent neural networks
- sequential monte carlo
- sample size
- active learning
- maximum likelihood