Multiplicative update rules for accelerating deep learning training and increasing robustness.
Manos KirtasNikolaos PassalisAnastasios TefasPublished in: Neurocomputing (2024)
Keyphrases
- deep learning
- update rules
- deep architectures
- update rule
- restricted boltzmann machine
- unsupervised learning
- machine learning
- mental models
- training set
- fixpoint
- weakly supervised
- ant colony optimization algorithm
- training examples
- loss function
- particle swarm optimisation
- image classification
- supervised learning
- image features