You are on page 1of 54

Neural Networks

CC524
Chapter 2
Basic Neuron Model, the Perceptron and the Learning
Algorithms (Perceptron & Gradient Learning)
2019-2020
The Basic Neuron Model
Gradient Descent Methods

1- Linear output Perceptron (LMS gradient descent


algorithm)
2- Gradient descent for Perceptron with nonlinear
output
The nonlinear output perceptron

1- The Sigmoid output


2- The Tanh output
3- The ReLu output (Rectified Linear Unit)

We should be able to calculate the gradient


learning equations for each case.

You might also like