Name: ___________________
Student
No: ___________________
Faculty: ___________________
Signature: ___________________
THE UNIVERSITY OF QUEENSLAND
FINAL
EXAMINATION, SEMESTER II, NOVEMBER 1999
IC230
LABORATORY INTRODUCTION TO NEURAL NETWORKS
Time:
Two (2) hours for working
Ten (10) minutes for perusal before examination begins
General Directions to Candidates
There are three sections to the paper, labeled
“Neural Networks”, “Symbolic Models” and “Genetic Algorithms”. You should
attempt to answer all three sections. The marks for each question are given in
brackets after each question (Total marks 50).
The examination has been set to take the full
two hours, so answer the questions that you know first, and only after
answering them, go back to the questions that you are less certain about.
Answer all questions on the examination paper.
If you need additional space use the back of the paper and clearly label each
answer.
SECTION A: NEURAL NETWORKS (30 MARKS)
Question 1
a) Draw a network of threshold logic units that implements the XOR function. Provide the values for all weights and biases. (2)
b)
What is the main difference
between the the networks necessary for the XOR
function as compared against the OR or AND functions?
Why is this difference necessary? (1).
Question 2
Name and describe the three main learning
paradigms. (2)
Question 3
a)
What are the defining
characteristics of an IAC network? (1).
b)
Provide the activation
equation of an IAC unit (1).
Question 4
The IAC Jets and Sharks network can be used to
demonstrate a number of important properties of neural information processing.
Define the following properties and describe how the Jets and Sharks network
can be used to demonstrate this property.
a)
Content addressability (1):
b)
Robustness to noise (1):
Question 5
a)
What is the word superiority
effect? (1)
b)
Describe briefly how the IAC
model of letter perception explains the word superiority effect? (1)
Question 6
a)
Define the term “distributed
representation”. (1)
b)
Provide two advantages of
distributed representations. (1)
Question 7
Suppose we have the following associations that
we wish to memorize using Hebbian learning:
Robin (1
1 0 0 0)
-> Flies (1 0)
Parrot (1 0 1 0 0)
-> Flies (1 0)
Kangaroo (0 0 0 0 1) -> Jumps (0 1)
a)
Calculate the memory matrix
to associate these patterns. (1)
b)
If we cue the memory with
the pattern for Emu (1 0 0
1 0) what output do we get and why? (1)
Question 8
a)
What is the difference
between synchronous and asynchronous updating in a Hopfield network? (1)
b)
Using the energy function:
Demonstrate
that in the Hopfield network containing a single pattern, that pattern is a
local minimum. (2).
Question 9
Use the perceptron
convergence procedure to find a set of weights for a three unit network that
implements the NOR function. (2)
NOR
function
Input 1 Input 2 Target
1 1 0
1 0 0
0 1 0
0 0 1
|
W1 |
W2 |
Bias |
I1 |
I2 |
Output |
Target |
|
0 |
-1 |
-0.5 |
1 |
1 |
|
0 |
|
|
|
|
1 |
0 |
|
0 |
|
|
|
|
0 |
1 |
|
0 |
|
|
|
|
0 |
0 |
|
1 |
|
|
|
|
1 |
1 |
|
0 |
|
|
|
|
1 |
0 |
|
0 |
|
|
|
|
0 |
1 |
|
0 |
|
|
|
|
0 |
0 |
|
1 |
Hint: The table shows how many steps it should
take to find (and verify) the solution.
Question 10
a)
Why is the sigmoid
activation function used in the backpropagation
algorithm rather than the threshold logic function? (1)
b)
Give the equation for
momentum. (1)
c)
What is the value of including
a momentum term in the backpropagation training
algorithm? (1).
Question 11
a)
Give the activation equation
for the self organizing map. (1)
b)
What is a twisted map and
how do they arise? (1).
Question 12
a)
What is the Stroop effect? (1)
b)
Describe how the Cohen,
Question 13
a)
In the field of human
memory, what is the difference between a retrieval task and a matching task?
(1)
b)
What is the difference
between an episodic task and a semantic task? (1)
c)
Give examples of each of the
different sorts of memory tasks in the following table: (1)
|
|
Matching Task |
Retrieval Task |
|
Episodic Task |
|
|
|
Semantic Task |
|
|
SECTION B: SYMBOLIC MODELS (10 MARKS)
Question 14
Soar is given a block stacking problem, starting with three blocks, A, B, and C. A is on top of B which is on the tabletop, and C is directly on the tabletop. Soar's goal is move the blocks one at a time to create a single stack of blocks, with A on top of B, and B on top of C. Using this problem as an example domain (3)
a) Explain what an
"impasse" is in Soar
b) Describe a possible impasse
situation
c) Explain how Soar might deal
with it.
Question
15
a) Under what conditions does
chunking occur in Soar? (1)
b) What are the benefits and
limitations of Soar's chunking procedure? (1)
Question
16
A possible
solution derived by Copycat to the problem abc : abd,
xyz : ? is wyz. This solution is driven by the fact that a is the first letter in the alphabet
and z is the last, allowing a
correspondence to be formed between them.
For the solution wyz
to be formed in the Workspace:
a) what would be the initial
and modified rule that allows this to occur ? (1)
initial rule: replace ________________________ by
_______________________
modified rule: replace
________________________ by ________________________
b) The slippages between the
concepts used in the initial and modified rule are either implied by or
incorporated in the correspondence that is built between the groups abc and xyz.
Denote what descriptions would be attached to the two groups, the slippages
that occur when the two groups are mapped, and what implied slippages are
required to derive the modified from the initial rule above. (1)
Implied conceptual slippage
c) In the following diagram
fill in the descriptions attached to the letters, the bonds that from between
the letters, and the correspondences between letters that are consistent with
your answer to (b) above. (1)
first
last
Question
17
What is the effect of clamping the temperature
(a) at a high value (i.e. 100)
(0.5)
(b) at a low value (i.e. 10)
(0.5)
(c) As the temperature falls
from a high to a low value during the formation of a solution, what do your
answers to (a) and (b) imply about the search strategy employed by
Copycat. (1)
SECTION C: GENETIC ALGORITHMS (10 MARKS)
Question 18
a)
What is the purpose of
"genetic operators" in a Genetic Algorithm? (1)
b)
Name two widely used genetic
operators and briefly describe the way in which each works. (4)
Question 19
a)
What is the "
b)
Under what circumstances
would you expect the