1. Fitness is a key measure when using evolutionary Al techniques. What does fitness refer to and why is it important? What is the fitness measure in the Travelling Salesman Problem? What was the fitness measure in Tierra?
2. Create a three-state Turing Machine rule set that works as an incrementer (adds one to a number). Remember that in binary 0 + 1 = 1 and 1 + 1 = 10. Assume the head is positioned over the first (most significant) bit of the number and that there are blanks on either side of the number. After adding 1, the head should end up positioned over the trailing blank where the program halts. For example if we begin with b1011b we should wind up with b1100b with the head over the rightmost blank
3. Back-propagation can be an important attribute of artificial neural networks What does back- propagation refer to and why is it important? Give an example of how this is used
4. Intelligent agents can be classified as software (web bots, spiders etc) and hardware (robots). Are there any differences in designing Al for a software agent as opposed to a hardware agent? Explain.
5. Alan Turing's thesis in his 1950 paper was "Can machines think?" So what do you think, can machines think? If not, will they ever be able to? Justify your answer.
6. The study of Al has moved towards studying emergence and emergent behaviour. What does emergence and emergent behaviour refer to? How does this relate to Al? Describe 2 examples of emergent behaviour.
7. Researchers in swarm technology study social insects, especially ants. Why? What's SO interesting about ant society? How can we use this? Give a couple of examples.
8. What are the differences between the Al necessary to operate an automated train and a self- driving car? Would you use an expert system, a fuzzy system, a genetic algorithm, genetic programming, ANN, or a swarm? Explain why this would be your choice
9. Claude Shannon's "Mouse", Ross Ashby's "Homeostat", and Grey Walters' Tortoise each illustrated different aspects of intelligent agents. What aspects were they? Where do we use the results of their research today?
10. When talking about mobile software agents, important issues are Transportation, Authentication, Secrecy, and Security. Choose two of these issues and describe what these issues refer to and explain why they are important.
Fitness is important because it defines the optimization methodology of the algorithm and therefore directly determines how effective it is. A fitness metric that aligns with the goal of the optimization will converge on a result quickly, while a poor fitness parameter will take substantially longer and could even not optimize correctly.
In the Travelling Salesman Problem, the fitness function is the length of the route—the shorter the route, the better.
In Tierra, there is no exogenous fitness function; the model is open-ended and therefore the computer processes are simply competing for survival.
2. Start in state A. Move to the right until a BLANK is observed, then transition to state B. In state B, move to the left, changing 1s to 0s until a 0 or BLANK is observed. Change it to a 1, then transition to state C. In state C, move to the right until a BLANK is found, upon which the state...
By purchasing this solution you'll be able to access the following files: