Experimental demonstration of a robust training method for strongly defective neuromorphic hardware.
William A. BordersAdvait MadhavanMatthew W. DanielsVasileia GeorgiouMartin Lueker-BodenTiffany S. SantosPatrick M. BragancaMark D. StilesJabez J. McClellandBrian D. HoskinsPublished in: CoRR (2023)
Keyphrases
- synthetic data
- significant improvement
- cost function
- training samples
- highly accurate
- error rate
- high accuracy
- input data
- training phase
- data sets
- training process
- high precision
- computationally efficient
- computational cost
- pairwise
- similarity measure
- feature set
- object detection
- detection method
- experimental evaluation
- computational complexity
- genetic algorithm
- neural network