- 340 Views
- Uploaded on

Download Presentation
## PowerPoint Slideshow about 'Artificial Neural Networks' - Faraday

**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.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript

### Artificial Neural Networks

Dan SimonCleveland State University

Neural Networks

- Artificial Neural Network (ANN): An information processing paradigm that is inspired by biological neurons
- Distinctive structure: Large number of simple, highly interconnected processing elements (neurons); parallel processing
- Inductive learning, that is, learning by example; an ANN is configured for a specific application through a learning process
- Learning involves adjustments to connections between the neurons

Inductive Learning

- Sometimes we can’t explain how we know something; we rely on our experience
- An ANN can generalize from expert knowledge and re-create expert behavior
- Example: An ER doctor considers a patient’s age, blood pressure, heart rate, ECG, etc., and makes an educated guess about whether or not the patient had a heart attack

The Birth of ANNs

- The first artificial neuron was proposed in 1943 by neurophysiologist Warren McCulloch and the psychologist/logician Walter Pitts
- No computing resources at that time

Feedforward ANN

How many hidden layers should we use? How many neurons should we use in each hidden layer?

Perceptrons

- A simple ANN introduced by Frank Rosenblatt in 1958
- Discredited by Marvin Minsky and Seymour Papert in 1969
- “Perceptrons have been widely publicized as 'pattern recognition' or 'learning machines' and as such have been discussed in a large number of books, journal articles, and voluminous 'reports'. Most of this writing ... is without scientific value …”

Perceptrons

x0=1

Three-dimensional single-layer perceptron

Problem: Given a set of training data (i.e., (x, y) pairs), find the weight vector {w} that correctly classifies the inputs.

w0

x1

w1

w2

x2

w3

x3

The Perceptron Training Rule

- t = target output, o = perceptron output
- Training rule: wi = e xi, where e = t o,and is the step size.Note that e = 0, 1 or 1.If e = 0, then don’t update the weight.If e = 1, then t = 1 and o = 0, so we need to increase wi if xi > 0, and decrease wi if xi < 0.Similar logic applies when e = 1.
- is often initialized to 0.1 and decreases as training progresses.

From Perceptrons to Backpropagation

- Perceptrons were dismissed because of:
- Limitations of single layer perceptrons
- The threshold function is not differentiable
- Multi-layer ANNs with differentiable activation functions allow much richer behaviors.

A multi-layer perceptron (MLP) is a feedforward ANN with at least one hidden layer.

Backpropagation

Derivative-based method for optimizing ANN weights.

1969: First described by Arthur Bryson and Yu-Chi Ho.

1970s-80s: Popularized by David Rumelhart, Geoffrey Hinton Ronald Williams, Paul Werbos; led to a renaissance in ANN research.

Derivative-based method for optimizing ANN weights

The Credit Assignment Problem

In a multi-layer ANN, how can we tell which weight should be varied to correct an output error? Answer: backpropagation.

Output 1

Wanted 0

Backpropagation

input neurons

hidden neurons

output neurons

x1

x1

v11

y1

w11

o1

a1

z1

v21

w21

v12

w12

x2

x2

v22

y2

w22

o2

a2

z2

Similar for a2 and y2

Similar for z2 and o2

tk = desired (target) value of k-th output neuron

no = number of output neurons

Sigmoid transfer function

D( j ) = {output neurons whose inputs come from the j-th middle-layer neuron}

vij aj yj

{ zk for all k D( j ) }

The Backpropagation Training Algorithm

- Randomly initialize weights {w} and {v}.
- Input sample xto get output o. Compute error E.
- Compute derivatives of E with respect to output weights {w} (two pages previous).
- Compute derivatives of E with respect to hidden weights {v} (previous page). Note that the results of step 3 are used for this computation; hence the term “backpropagation”).
- Repeat step 4 for additional hidden layers as needed.
- Use gradient descent to update weights {w} and {v}. Go to step 2 for the next sample/iteration.

XOR Example

x2

Not linearly separable. This is a very simple problem, but early ANNs were unable to solve it.

y = sign(x1x2)

1

0

x1

0

1

XOR Example

x1

x1

v11

y1

w11

o1

a1

z1

Bias nodes at both the input and hidden layer

w21

v12

v21

y2

w31

x2

x2

v22

a2

v31

v32

1

1

1

1

Backprop.m

XOR Example

x2

Homework: Record the weights for the trained ANN, input various (x1, x2) combinations to the ANN to see how well it can generalize.

1

0

x1

0

1

Backpropagation Issues

- Momentum: wij wij – jyi + wij,previousWhat value of should we use?
- Backpropagation is a local optimizer
- Combine it with a global optimizer (e.g., BBO)
- Run backprop with multiple initial conditions
- Add random noise to input data and/or weights to improve generalization

Backpropagation Issues

Batch backpropagation

Don’t forget to adjust the learning rate!

- Randomly initialize weights {w} and {v}
- While not (termination criteria)
- For i = 1 to (number of training samples)
- Input sample xito get output oi. Compute error Ei
- Compute dEi / dw and dEi / dv
- Next sample
- dE / dw = dEi / dw and dE / dv = dEi / dv
- Use gradient descent to update weights {w} and {v}.
- End while

Backpropagation Issues

Weight decay

- wij wij – jyi – dwijThis tends to decrease weight magnitudes unless they are reinforced with backpropd 0.001
- This corresponds to adding a term to the error function that penalizes the weight magnitudes

Backpropagation Issues

Quickprop (Scott Fahlman, 1988)

- Backpropagation is notoriously slow.
- Quickprop has the same philosophy as Newton-Raphson.Assume the error surface is quadratic and jump in one step to the minimum of the quadratic.

Backpropagation Issues

- Other activation functions
- Sigmoid: f(x) = (1+e–x)–1
- Hyperbolic tangent: f(x) = tanh(x)
- Step: f(x) = U(x)
- Tan Sigmoid: f(x) = (ecx – e–cx) / (ecx + e–cx) for some positive constant c
- How many hidden layers should we use?

Universal Approximation Theorem

- A feed-forward ANN with one hidden layer and a finite number of neurons can approximate any continuous function to any desired accuracy.
- The ANN activation functions can be any continuous, nonconstant, bounded, monotonically increasing functions.
- The desired weights may not be obtainable via backpropagation.
- George Cybenko, 1989; Kurt Hornik, 1991

Termination Criterion

If we train too long we begin to “memorize” the training data and lose the ability to generalize.

Train with a validation/test set.

Error

Validation/Test Set

Training Set

Termination Criterion

Cross Validation

- N data partitions
- N training runs, each using (N1) partitions for training and 1 partition for validation/test
- Each training run, store number of epochs cifor the best test set performance (i=1,…,N)
- cave = mean{ci}
- Train on all data for cave epochs

Adaptive Backpropagation

Recall standard weight update: wij wij – jyi

- With adaptive learning rates, each weight wij has its own rate ij
- If the sign of wij is the same over several backprop updates, then increase ij
- If the sign of wij is not the same over several backprop updates, then decrease ij

Double Backpropagation

P = number of input training patterns. We want an ANN that can generalize. So input changes should not result in large error changes.

In addition to minimizing the training error:

Also minimize the sensitivity of training error to input data:

Other ANN Training Methods

Gradient-free approaches (GAs, BBO, etc.)

- Global optimization
- Combination with gradient descent
- We can train the structure as well as the weights
- We can use non-differentiable activation functions
- We can use non-differentiable cost functions

BBO.m

Classification Benchmarks

The Iris classification problem

- 150 data samples
- Four input feature values (sepal length and width, and petal length and width)
- Three types of irises: Setosa, Versicolour, and Virginica

Classification Benchmarks

- The two-spirals classification problem
- UC Irvine Machine Learning Repository – http://archive.ics.uci.edu/ml194 benchmarks!

Radial Basis Functions

J. Moody and C. Darken, 1989

Universal approximators

N middle-layer neurons

Inputs x

Activation functions f (x, ci)

Output weights wik

yk = wikf (x, ci)

= wik ( ||xci|| )

(.) is a basis function

limx ( ||xci|| ) = 0

{ ci } are the N RBF centers

Radial Basis Functions

Common basis functions:

- Gaussian: ( ||xci|| ) = exp(||xci||2 / 2) is the width of the basis function
- Many other proposed basis functions

Radial Basis Functions

Suppose we have the data set (xi, yi), i = 1, …, N

Each xi is multidimensional, each yi is scalar

Set ci = xi, i = 1, …, N

Define gik = ( || xi xk|| )

Input each xi to the RBF to obtain:

Gw = y

G is nonsingular if {xi} are distinct

Solve for w

Global minimum (assuming fixed c and )

Radial Basis Functions

We again have the data set (xi, yi), i = 1, …, N

Each xi is multidimensional, each yi is scalar

ck are given for (k = 1, …, m), and m < N

Define gik = ( || xi ck|| )

Input each xi to the RBF to obtain:

Gw = y

w = (GTG)1GT = G+y

Radial Basis Functions

How can we choose the RBF centers?

- Randomly select them from the inputs
- Use a clustering algorithm
- Other options (BBO?)

How can we choose the RBF widths?

Other Types of ANNs

Many other types of ANNs

- Cerebellar Model Articulation Controller (CMAC)
- Spiking neural networks
- Self-organizing map (SOM)
- Recurrent neural network (RNN)
- Hopfield network
- Boltzman machine
- Cascade-Correlation
- and many others …

Sources

- Neural Networks, by C. Stergiou and D. Siganos, www.doc.ic.ac.uk/~nd/ surprise_96/journal/vol4/cs11/report.html
- The Backpropagation Algorithm, by A. Venkataraman, www.speech.sri.com/people/anand/771/html/node37.html
- CS 478 Course Notes, by Tony Martinez, http://axon.cs.byu.edu/~martinez

Download Presentation

Connecting to Server..