The fundamental component of all DL models is an artificial neuron. The artificial neuron is inspired by the working of biological neurons. It consists of some inputs connected via weights (also called synaptic connections), the weighted sum of all the inputs goes through a processing function (called the activation function) and generates a non-linear output.
The following screenshot shows A biological Neuron and An Artificial Neuron:
A biological neuron and an artificial neuron
If Xi is the ith input to the artificial neuron (j) connected via the synaptic connection wij, then, the net input to the neuron, commonly called ...
