**Artificial Neurons and its activation function**

Just like bio-logical neuron cell, the artificial neuron
contains several inputs, output terminals and a computation unit (nucleus) at
the center, as shown in the below figure. Each input connected to the neuron
cell has its own weight represents the strength of the input signal.

Artificial Neuron |

X1, X2, X3, …. Xn are the inputs of artificial neuron and
these are multiplied with corresponding weights (W1, W2, W3, … Wn) before
entering in to the nucleus. The nucleus will sum up all these weighted input
values which produce “activation”.

The activation is given as

Where Wj is the weight associated with the jth input (Xj).

McCulloch and Pitts were proposed a binary threshold unit as
a computational model for an artificial neuron cell.

i.e., if the sum of the weighted inputs of the
neuron cell is greater than threshold value gives a output ‘1’ otherwise is
gives ‘0’ as output.McCulloch-Pitts model of a artificial neuron cell |

## 0 comments :

## Post a Comment

Thanks for visiting NPEDUCATIONS. We will contact you soon