A heuristic approach to the hyperparameters in training spiking neural networks using spike-timing-dependent plasticity.
Dawid PolapMarcin WozniakWaldemar HolubowskiRobertas DamaseviciusPublished in: Neural Comput. Appl. (2022)
Keyphrases
- spiking neural networks
- hyperparameters
- training algorithm
- biologically inspired
- model selection
- cross validation
- closed form
- bayesian framework
- feed forward
- bayesian inference
- biologically plausible
- support vector
- artificial neural networks
- prior information
- random sampling
- learning rules
- sample size
- training set
- em algorithm
- noise level
- spiking neurons
- neural network
- incremental learning
- incomplete data
- missing values
- back propagation
- maximum likelihood
- training process
- radial basis function
- maximum a posteriori
- supervised learning
- probabilistic model
- pairwise
- parameter settings
- generative model
- prior knowledge