Don't stop the training: continuously-updating self-supervised algorithms best account for auditory responses in the cortex.
Pierre OrhanYves BoubenecJean-Rémi KingPublished in: CoRR (2022)
Keyphrases
- learning algorithm
- real time
- orders of magnitude
- computationally efficient
- computational cost
- early stopping
- recently developed
- times faster
- computational efficiency
- information processing
- signal processing
- training set
- objective function
- theoretical analysis
- computational complexity
- data structure
- convergence rate
- neural network training
- neural network