Loading in 5 sec....

Derivation of a Learning Rule for Perceptrons PowerPoint Presentation

Derivation of a Learning Rule for Perceptrons

- By
**uma** - Follow User

- 102 Views
- Uploaded on

Download Presentation
## PowerPoint Slideshow about ' Derivation of a Learning Rule for Perceptrons ' - uma

**An Image/Link below is provided (as is) to download presentation**
Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

Presentation Transcript

Single Layer Perceptrons

x1

wk1

x2

wk2

.

.

.

wkm

xm

Derivation of a Learning Rule for PerceptronsAdaline

(Adaptive Linear Element)

Widrow [1962]

Goal:

Single Layer Perceptrons

Least Mean Squares (LMS)- The following cost function (error function) should be minimized:

i : index of data set, the ith data set

j : index of input, the jth input

Single Layer Perceptrons

Adaline Learning Rule- With

then

- As already obtained before,

Weight Modification Rule

- Defining

we can write

Single Layer Perceptrons

Adaline Learning Modes- Batch Learning Mode

- Incremental Learning Mode

Single Layer Perceptrons

Tangent Sigmoid Activation Functionx1

wk1

x2

wk2

.

.

.

wkm

xm

Goal:

Single Layer Perceptrons

Logarithmic Sigmoid Activation Functionx1

wk1

x2

wk2

.

.

.

wkm

xm

Goal:

Single Layer Perceptrons

Derivation of Learning Rules- For arbitrary activation function,

Single Layer Perceptrons

Derivation of Learning RulesDepends on the activation function used

Single Layer Perceptrons

Derivation of Learning RulesLinear function

Tangent sigmoid

function

Logarithmic sigmoid

function

Single Layer Perceptrons

x1

w11

x2

w12

Homework 3Given a neuron with linear activation function (a=0.5), write an m-file that will calculate the weights w11 and w12 so that the input [x1;x2] can match output y1 the best.

- Use initial values w11=1 and w12=1.5, and η= 0.01.
- Determine the required number of iterations.
- Note: Submit the m-file in hardcopy and softcopy.

[x1;x2]=[2;3]

[x1;x2]=[[2 1];[3 1]]

Case 2

Case 1

[y1]=[5 2]

[y1]=[5]

- Odd-numbered Student ID

- Even-numbered Student ID

Single Layer Perceptrons

x1

w11

x2

w12

Homework 3AGiven a neuron with a certain activation function, write an m-file that will calculate the weights w11 and w12 so that the input [x1;x2] can match output y1 the best.

- Use initial values w11=0.5 and w12=–0.5, and η= 0.01.
- Determine the required number of iterations.
- Note: Submit the m-file in hardcopy and softcopy.

[x1]=[0.2 0.5 0.4]

[x2]=[0.5 0.8 0.3]

[y1]=[0.1 0.7 0.9]

?

- Even Student ID:Tangent sigmoid function

- Odd Student ID:Logarithmic sigmoid function

Multi Layer Perceptrons

x1

x2

x3

wlk

wji

wkj

MLP ArchitectureHidden layers

Input

layer

Output

layer

y1

Outputs

Inputs

y2

- Possessessigmoid activation functionsin the neurons to enable modeling of nonlinearity.
- Contains one or more “hidden layers”.
- Trained using the “Backpropagation” algorithm.

Multi Layer Perceptrons

MLP Design Consideration- What activation functions should be used?
- How many inputs does the network need?
- How many hidden layers does the network need?
- How many hidden neurons per hidden layer?
- How many outputs should the network have?

- There is no standard methodology to determine these values. Even there is some heuristic points, final values are determinate by a trial and error procedure.

Multi Layer Perceptrons

x1

x2

x3

wlk

wji

wkj

Advantages of MLP- MLP with one hidden layer is a universal approximator.
- MLP can approximate any function within any preset accuracy
- The conditions: the weights and the biases are appropriately assigned through the use of adequate learning algorithm.

- MLP can be applied directly in identification and control of dynamic system with nonlinear relationship between input and output.
- MLP delivers the best compromise between number of parameters, structure complexity, and calculation cost.

Multi Layer Perceptrons

f(.)

f(.)

f(.)

Learning Algorithm of MLPFunction signal

Error signal

- Computations at each neuron j:
- Neuron output, yj
- Vector of error gradient, ¶E/¶wji

Forward propagation

“Backpropagation

Learning Algorithm”

Backward propagation

Multi Layer Perceptrons

Backpropagation Learning AlgorithmIf node j is an output node,

dj(n)

yj(n)

netj(n)

wji(n)

ej(n)

yi(n)

-1

f(.)

Multi Layer Perceptrons

Backpropagation Learning AlgorithmIf node j is a hidden node,

dk(n)

netk(n)

yj(n)

yk(n)

netj(n)

wji(n)

wkj(n)

yi(n)

ek(n)

f(.)

f(.)

-1

Multi Layer Perceptrons

k

j

i

Right

Left

k

j

i

Right

Left

MLP Training- Forward Pass
- Fix wji(n)
- Compute yj(n)

- Backward Pass
- Calculate dj(n)
- Update weights wji(n+1)

Download Presentation

Connecting to Server..