Backend class

The Backend class is the base class for backend implementations. Each backend is implemented as a library, allowing you to:

• Run training and execution of the spiking neural network on various devices
  

  A neural network based on artificial neurons that perform temporal summation of input signals represented by spikes and weighted according to the weights of input connections, and generate an output pulse according to a certain threshold-based activation function.

• Read output spikes received from the computer
  

  Short-term action potential generated by a neuron in a population as a result of the neuron attribute changes.

• Get internal data to monitor the neural network operation
• Upload internal data of the neural network operation monitoring received from the computer

To initialize the backend, first download projections containing synapses and populations containing neurons. During initialization of the backend, the loaded synapse projections are subscribed to the spikes of the associated populations or input channels, and the downloaded populations are subscribed to synaptic impacts of the associated projections.

Each object of the Backend class has the following attributes:

• Basic backend data.
• Message bus. Using the message bus, the backend ensures the message exchange between the objects of the core library, as well as sending messages from modifiers to neurons and synapses.
  

  Code that modifies the behavior of a neuron or synapse, for example, adds synaptic plasticity.

• Message bus endpoint.
• Backend initialization status.
• Neural network execution status on the backend.
• Backend devices.
• Neural network execution step number.

Page top