Research Article
Intralayer-Connected Spiking Neural Network with Hybrid Training Using Backpropagation and Probabilistic Spike-Timing Dependent Plasticity
| | Require: Spike probability at each time point , number of layers , number of time points , and weight matrix parameters | | | Ensure: Update the weight | | (1) | for to do | | (2) | if layer is not fully connected then | | (3) | Continue | | (4) | end if | | (5) | Matrix | | (6) | forto − 1 do | | (7) | fortodo | | (8) | Calculate using equation (19) | | (9) | Calculate using equation (19) | | (10) | | | (11) | Calculate the update weight matrix between And | | (12) | Calculate using equation (16) | | (13) | | | (14) | end for | | (15) | end for | | (16) | Update the connection weight for layer | | (17) | | | (18) | end for |
|
