Title: Neural%20Networks%20Multilayer%20Perceptron%20(MLP)%20Oscar%20Herrera%20Alc
1Neural NetworksMultilayer Perceptron
(MLP)Oscar Herrera Alcántaraheoscar_at_yahoo.com
2Outline
- Neuron
- Artificial neural networks
- Activation functions
- Perceptrons
- Multilayer perceptrons
- Backpropagation
- Generalization
3Neuron
- A neuron is a cell in the brain
- collection, processing, and dissemination of
electrical signals - neurons of gt 20 types, synapses,
1ms-10ms cycle time - brains information processing relies on networks
of such neurons
4Biological Motivation
- dendrites nerve fibres carrying electrical
signals to the cell - cell body computes a non-linear function of its
inputs - axon single long fiber that carries the
electrical signal from the cell body to other
neurons - synapse the point of contact between the axon of
one cell and the dendrite of another, regulating
a chemical connection whose strength affects the
input to the cell.
5Artificial neural networks
- A mathematical model of the neuron is
McCulloch-Pitts unit - Neural networks consists of nodes (units)
connected by directed links
1 x1 x2 x3 xm
b Bias
wi1
Neuron i
v
S
y
j
Wim
Synaptic Weights
Induced local field
Activation
Inputs
Output
Activation potential
function
- A bias weight Wi,0 connected to a fixed input
xi,0 1
6Activation functions
- Step function or Threshold function
- Sigmoid function
- Hyperbolic tangent function
7Perceptron learning
- Learn by adjusting weights to reduce error on
training set - Error correction learning rule
- Perform optimization search by gradient descent
8Implementing logic functions
- McCulloch-Pitts unit can implement any Boolean
function
9Expressiveness of perceptrons
- A perceptron
- can represent AND, OR, NOT
- can represent a linear separator (function) in
input space
10Multilayer Perceptron (MLP) Architecture
Bias
11Solve XOR problem using MLPs
- A two-layer network with two nodes in the hidden
layer - The hidden layer maps the points from non linear
separable space to linear separable space. - The output layer finds a decision line
j (v)
12Back-propagation Algorithm
1. Initialization. Weights are initialized with
random values whose mean is zero 2.
Presentations of training examples 3. Forward
computation 4.-Backward computation for the
neuron j of the hidden layer l for the
neuron j of the output layer L
5.- Iteration. Repeat step 2 to 4 until Elt
desired error a the momentum parameter is
ajusted h the learning-rate parameter is
ajusted
13MLP Training
k
j
i
Right
Left
- Forward Pass
- Fix wji(n)
- Compute yj(n)
- Backward Pass
- Calculate dj(n)
- Update weights wji(n1)
y
x
k
j
i
Right
Left
14Generalization
- Total Data are divided in two parts
- Data Training (80)
- MLP is trained with Data Training
- Data Test (20)
- MLP is tested with Data Test
- Generalization
- MLP is used with inputs which have never been
presented in order to predict the outputs