Neural networks (Norwegian: Nevrale nett) are either biological neural networks consisting of neurons, or in a modern and simplified sense artificial neural networks consisting of artificial nodes with activation functions. The connections of the biological neuron are modeled as weights in the artificial node. Positive weights reflects excitatory synapses, while negative weights reflects inhibitory synapses.
Posts
- training with autoencoders – unsupervised learning and autoencoders
- Strong EPSP – strong postsynaptic response and multimodality
- Minicolumns as a language – the strange similarities with formal semantics
- Rate encoded location – place encode like a brain