Transcript Artificial Neural Networks

Rohit Ray
ESE 251
What are Artificial Neural
Networks?
 ANN are inspired by models of the biological nervous




systems such as the brain
Novel structure by which to process information
Number of highly interconnected processing elements
(neurons) working in unison to solve specific
problems.
Recent Development
First artificial neuron -1943 by Warren McCulloch and
Walter Pits.
 But the technology available at that time did not allow
them to do too much.
Biological Inspiration
 Animals are able to react adaptively to changes in their
external and internal environment, and they use their
nervous system to perform these behaviours.
 An appropriate model/simulation of the nervous system
should be able to produce similar responses and behaviours
in artificial systems.
 The nervous system is build by relatively simple units, the
neurons, so copying their behaviour and functionality
should be the solution.
From http://www.scienceclarified.com/scitech/ArtificialIntelligence/Mind-Versus-Metal.html
Artificial Neural Networks (ANNs),
 Work in the same way as the brain's neural network.
 An artificial neuron has a number of connections or inputs.
 It is based on the belief that the way the brain works is all
about making the right connections
 Are good for prediction and estimation when:
 Inputs are well understood
 Output is well understood
From http://www.scienceclarified.com/scitech/ArtificialIntelligence/Mind-Versus-Metal.html
Artificial Neuron
From
http://www.uic.edu/classes/idsc/ids572cna/Neural%20Networks_1
.pdf
Example of a ANN
How does it work
 Neural Network Training
 Training - process of setting the best weights on the
edges connecting all the units in the network
 Use the training set to calculate weights such that ANN
output is as close as possible to the desired output for as
many of the examples in the training set as possible
From
http://www.uic.edu/classes/idsc/ids572cna/Neural%20Networks_1
.pdf
Training an ANN
 Adjust weights such that the application of inputs produce desired
 outputs (as close as possible)
 Input data is continuously applied, actual outputs calculated, and
weights are adjusted
 Weights should converge to some value after many rounds of training
 Supervised training
 Adjust weights such that differences between desired and actual outputs
are minimized
 Desired output: dependent variable in training data
 Each training example specifies {independent variables, dependent
variable}
 Unsupervised training
 No dependent variable specified in training data
 Train the NN such that similar input data should generate same output
http://www.scienceclarified.com/scitech/ArtificialIntelligence/Mind-Versus-Metal.html#ixzz0V6k2i38h
Example: Will the teacher give a quiz?

To help solve this question a programmer is provided with the following options
 The teacher loves giving quizzes = 0.2.
 The teacher has not given a quiz in two weeks = 0.1.
 The teacher gave the last three quizzes on Fridays = 0.3.

The sum of the input weights equals 0.6.

The threshold assigned to that neuron is 0.5. In this case, the net value of the neuron exceeds the
threshold number so the artificial neuron is fired. This process occurs again and again in rapid
succession until the process is completed.

If the ANN is wrong, and the teacher does not give a quiz on Friday, then the weights are lowered.
Each time a correct connection is made, the weight is increased. The next time the question is asked,
the ANN will be more likely to answer correctly.



The proper connections are weighted so that there is more chance that the machine will choose that
connection the next time. After hundreds of repeated training processes, the correct neural network
connections are strengthened and remembered, just like a memory in the human brain
A computer can make millions of trial-and-error attempts at lightning speed.
Comparison to other methods
 Simulated Annealing
 More accurate results
 Much slower
 Genetic Algorithms
 More accurate results
 Slower
Application of ANNs
 Broad applicability to real world business problems.
 Since neural networks are best at identifying patterns
or trends in data, they are well suited for prediction or
forecasting needs including:
 sales forecasting
 industrial process control
 customer research
 data validation
 Risk management
 target marketing
Application Cont.
 Medicine



Recognizing diseases from various scans
 no need to provide a specific algorithm on how to identify the
disease
Modeling Parts of the Human body
 cardiovascular system must mimic the relationship among
physiological variables (i.e., heart rate, systolic and diastolic
blood pressures, and breathing rate)
 specific to an individual (physical condition)
Instant Physician(1980’s)
 Given a set of symptoms it will then find the full stored pattern
that represents the "best" diagnosis and treatment.
Conclusion
 Computing world lots to gain from ANNs
 Ability to learn by example makes them very flexible and
powerful
 no need to devise an algorithm in order to perform a
specific task; i.e. there is no need to understand the
internal mechanisms of that task
 Regularly used in medicine and business
 Used to make models
 Find optimums, recognize patterns
Works Cited
 http://www.scienceclarified.com/scitech/Artificial-
Intelligence/Mind-Versus-Metal.html
 http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol
4/cs11/report.html
 http://www.uic.edu/classes/idsc/ids572cna/Neural
%20Networks_1.pdf