Activation Function for a Bayesian Classifier

Assignment of Nodes

The Output nodes are assigned such that each node represents one particular class in the classification problem.

Softmax (Normalized Exponential)

$$y_i=\frac{e^{v_i}}{{\sum }_{j=1}^N{e}^{v_j}}\in [0,1]$$

Where,

• $v_i\to$ It is the output from the Hidden Layer before applying the activation function.