presented by scott lichtor an introduction to neural networks
TRANSCRIPT
Overview• Basics of the Nervous System
– Neurons– Synapses– Action Potentials
• Neural Networks– Abstract Neurons– More Complicated Neurons
– Learning– Supervised– Unsupervised– Reinforcement
• Conclusion
Basics of the Nervous SystemThe nervous system coordinates the
actions of an animalBody parts send messages to the brainBrain sends messages to body partsThe basic unit of the nervous system is the
neuron
NeuronsReceive messages at the dendritesMessage is sent quickly down the axon using
electrical impulsesWhat happens when the signal reaches the end
of the axon?
Image taken from img460.imageshack.us
SynapsesChemical Synapses
SlowStrongCan be transmitted over long distances
Image taken from http://www.airlinesafety.com/editorials
Action PotentialsAction potentials are shocks to a particular
neuronThe shock travels along the affected
neuronThen, the action potential is transmitted
from the affected neuron to the neurons connected to it
The shock is transmitted to its destination in the same fashion
Abstract NeuronsSo biological neurons can be used to send
modified messages from place to placeCan be used to accomplish very complex
tasks using relatively simple partsCan neurons represent other things/be
used for other objectives?
Abstract NeuronsNeurons can represent neuron-like thingsInputs -> Processes -> Outputs
Image taken from http://3.bp.blogspot.com/
Abstract NeuronsCan “train” the neurons
Neurons fire (output 1) under certain patterns
Don’t fire (output 0) under other patternsFiring rule: if an outcome doesn’t fit in either
pattern, it fires if it has more in common with the first set, and doesn’t fire if it has more in common with the second set.
If there’s a tie, the neuron may fire, or it may not
Abstract Neurons Example
A neuron takes three inputs (X1, X2, X3)The neuron is trained to output 1 if the inputs are 111
or 101Trained to output 0 if the inputs are 000 or 001Before firing rule:
After firing rule:
X1 0 0 0 0 1 1 1 1
X2 0 0 1 1 0 0 1 1
X3 0 1 0 1 0 1 0 1
Out 0 0 0/1 0/1 0/1 1 0/1 1
X1 0 0 0 0 1 1 1 1
X2 0 0 1 1 0 0 1 1
X3 0 1 0 1 0 1 0 1
Out 0 0 0 0/1 0/1 1 1 1
Abstract NeuronsThe abstract neuron model can be used for
pattern recognitionExample: determine whether a ‘T’ or ‘H’ is
displayedCan we model more complicated processes
with neurons?
More Complicated NeuronsMcCulloch and Pitts modelDifference from previous model: inputs are
weighted.Add weighted inputs together: if the sum is
greater than a threshold, then the neuron fires
Image taken from http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html
More Complicated NeuronsNew model is very adaptable/powerfulInput weights and threshold can be
changed so the neuron responds differently/more accurately to a situation
Pavlov’s dogVarious algorithms adapt neurons and
neural networks to situationsDelta rule (feed-forward networks)Back-error projection (feedback networks)
LearningFor the network to adapt, it must learn.There are three types of learning used with
neural networks:Supervised learningUnsupervised learningReinforcement learning
Supervised LearningIn supervised learning, the system learns using
test data given from an external teacherThe test data tells the system what outputs
result from certain inputsThe system tries to match the response of the
test data, i.e. minimize the error between the neural network outputs and the test data outputs given the same inputs
Image taken from http://www.learnartificialneuralnetworks.com
Unsupervised LearningIn unsupervised learning, the network is
given no output dataInstead, the network is given just input
dataThe goal of the network, then, is to group
the input dataExample: mortgage requests
The network is given credit ratings, size of mortgage, interest rate, etc.
The network groups the data; probably into accept and deny
Reinforcement LearningNetwork performs actions on the input dataThe environment grades the network (good
or bad)The network makes adjustments
accordinglyMiddle ground between supervised and
unsupervised learning
ConclusionThe learning aspect of neural networks
makes their applications astoundingFor computers, one has to know how to
solve a particular problemNeural networks can solve problems that
one doesn’t know how to solve
ConclusionJust some of the uses: sales forecasting,
stock market prediction, customer research, modeling and diagnosing the cardiovascular system, “Instant Physician”, interpretation of multi-meaning Chinese words, facial recognition, etc. etc. etc.
Something I found interesting: the interconnectedness of different subjects
Sourceshttp://www.doc.ic.ac.uk/~nd/surprise_96/
journal/vol4/cs11/report.htmlhttp://www.learnartificialneuralnetworks.com/http://www.ryerson.ca/~dgrimsha/courses/
cps721/unsupervised.htmlhttp://www.willamette.edu/~gorr/classes/
cs449/intro.htmlhttp://www.statsoft.com/textbook/
stneunet.htmlhttp://www.wikipedia.org