1

ARTIFICIAL NEURAL NETWORKS

End of Semester Presentation

Presented by:Saif Al Kalbani 39579/12

20-05-2014

ECCE6206Switching Theory: Design and PracticeSpring 2014

Sultan Qaboos UniversityCollege of Engineering

Department of Electrical and Computer Engineering

2

Outline

Introduction General Architecture Learning Examples Applications Neuro-Fuzzy Conclusion

3

Applications

Input is high-dimensional discrete or real-valued (e.g. raw sensor input)

Output is discrete or real valued Output is a vector of values Form of target function is unknown Control System

Transfer function with huge number of inputs

Unknown transfer function

4

General Architecture

Network Interconnections Layers

Input layers Hidden layers Output layer

Inp

uts

Output

5

General Architecture

6

General Architecture

Threshold switching units Weighted interconnections among units Highly parallel, distributed processing Learning by tuning the connection weights

7

General Architecture

8

General Architecture

• Layers• Activation function• Learning

9

Layers

• The input layer.– Introduces input values into the network.– No activation function or other processing.

• The hidden layer(s).– Perform classification of features– Two hidden layers are sufficient to solve any

problem– Features imply more layers may be better

• The output layer.– Functionally just like the hidden layers– Outputs are passed on to the world outside the

neural network.

10

Examples

11

Activation Function

12

Learning

• Adjust neural network weights to map inputs to outputs.

• Use a set of sample patterns where the desired output (given the inputs presented) is known.

• The purpose is to learn to generalize– Recognize features which are common to

good and bad exemplars– Types

– Supervises– Unsupervised

13

Supervised Learning

Training and test data sets Training set; input & target

14

Learning

wi = wi + wi

wi = (t - o) xi

t=c(x) is the target value o is the perceptron output Is a small constant (e.g. 0.1) called learning rate

• If the output is correct (t=o) the weights wi are not changed

• If the output is incorrect (to) the weights wi are changed such that the output of the perceptron for the new weights is closer to t.

The algorithm converges to the correct classification• if the training data is linearly separable and is sufficiently

small

15

Learning

For ANDA B Output0 0 00 1 01 0 01 1 1

t = 0.15

y

x

W = 0.0

W = 0.0

x y Summation Output

0 0 (0*0.0) + (0*0.0) = 0.0 0

0 1 (0*0.0) + (1*0.0) = 0.0 0

1 0 (1*0.0) + (0*0.0) = 0.0 0

1 1 (1*0.0) + (1*0.0) = 0.0 0

T-o=1wi = (t - o) xi

=0.1wi=(1)*1*0.1=0.1Then Add 0.1 to the weights

16

Learning

For ANDA B Output0 0 00 1 01 0 01 1 1

t = 0.15

y

x

W = 0.1

W = 0.1

x y Summation Output

0 0 (0*0.1) + (0*0.1) = 0.0 0

0 1 (0*0.1) + (1*0.1) = 0.1 0

1 0 (1*0.1) + (0*0.1) = 0.1 0

1 1 (1*0.1) + (1*0.1) = 0.2 1

17

Decision Boundary

X1

X2

A

B

A

A

AA

AA

B

B

B

B

B

B

B DecisionBoundary

18

Strength

Solving complex problems Inputs are complex, large, unknown Transfer function is unknown

Adaptation Adaptive controllers Learning process

19

Shortfalls

Learning Weights

Processing time Delays in processing Sensing

Set up Layers

20

Application Example

• Engine Control Unit (ECU) in new cars• Fuel injector• The behaviour of a car engine is influenced

by a large number of parameters– temperature at various points– fuel/air mixture– lubricant viscosity.

• Major companies have used neural networks to dynamically tune an engine depending on current settings.

21

Neuro-Fuzzy

Hybrid controllers ANN controllers Fuzzy logic Controllers

Adaptation Rules

Membership functions

22

Neuro-Fuzzy

23

Conclusion

Ability of ANN to Adapt through learning Solve complex systems

ANN is claimed to be able to solve any problem with a maximum of two hidden layers

24

Thank You

Q&A

25 Back-up

26

Fuzzy System