Minimize

loss

Normalize

output

Introduce

non-linearity

Control

learning rate

Sigmoid

Linear

function

Step

function

None

Hidden

weights

Activation

Function

Output

layer

Bias

Input,

Output, Activation Function

Weights,

Bias, Output

Threshold,

Signal, Response

Dendrites,

Axon, Soma

Fail

to converge

Converge

Always

classify correctly

None

Supervised

extension of SOM

Genetic

network

Unsupervised

version of Adaline

None

f’(x ) =λ.f(x) [1-

f(x)]

f’(x ) =λ/2 [1+

f(x)][1-f(x)]

f’(x ) =λ.f(x) [1+ f(x)]

None

D

A

C

B

D

A

C

B

Gaussian

function

Sigmoid

function

Linear

function

Step

function