All Matches

Solution Library

Expert Answer

Textbooks

Search Textbook questions, tutors and Books

Oops, something went wrong!

Change your search query and then try again

Result Not Found

Books FREE
Tutors
Study Help
Hire a Tutor
AI Study Help New

Search
Sign In
Register

study help
business
management and artificial intelligence

Questions and Answers of Management And Artificial Intelligence

6. Write a program for a GA-based tic-tac-toe player.
5. Write a program to have a GA solve the 15-puzzle. The input to your program is a random arrangement of the tiles. The output is the tiles in order or a message that a solution is not possible
4. Write a program to have a GA determine the chromatic number of a graph (see Exercise 4).Test your program on the graphs depicted in Figures 2.39 and 2.40.
3. Write a program to have a GA solve the Missionaries and Cannibals problem (see Exercise 3).
2. Write a program to have a GA solve the 4-Queens problem (see Exercise 2).
1. Write a program that uses Monte Carlo simulation to approximate the value of π. (See Exercise 1 above). Use pairs of random numbers over [0, 1) instead of darts.
9. Until Darwin’s theory of evolution, people believed that living systems were designed by God (or a God-like figure). William Paley was a theologian who proposed in the 1802 book Natural Theology
8. Design a GP to determine the chromatic number of a graph (see Exercise 12.3). You will need to use functions thata. Assign a color to a node.b. Change the color of a node when necessary.c. Keep
7. How would you design a GA-based strategy for the iterated Prisoner’s Dilemma (Chapter 4)?
6. How would you formulate a GA that is capable of playing tic-tac-toe?
5. Design a GA solution for the 15-puzzle.
4. Design a GA solution for determining the chromatic number of a graph (Chapter 2). How does your fitness function avoid infeasible solutions? How does it reward solutions using fewer colors?
3. Design a GA solution for the Missionaries and Cannibals problem (Chapter 2). How does your fitness function measure closeness to the goal? How does it prevent unsafe states from occurring?
2. Design a GA solution for the 4-Queens problem (Chapter 2). Be sure to specify your representation and your fitness function.
1. Drawn on the dartboard is ¼ of a circle as shown. You throw 100 darts at the board. Assume that all darts land randomly somewhere on the board.How can you use this experiment to approximate the
18. Cite several future applications for swarm robotics at both the macroscopic and microscopic level.
17. What possible applications do you envision for the cemetery formation example?
16. Observe that in Figure 12.32, some ants do not follow the shortest path from the nest to a food source. What useful purposes do these supposedly misguided foragers play?
15. Explain the role that pheromone evaporation plays in the shortest path example explained in Section 5.
14. What is stigmergy? Why is this a useful means of communication?
13. List three aspiration criteria in tabu search and explain why they are helpful.
12.a. In tabu search, do tabu lists encourage exploitation or exploration?b. Same question for aspiration lists?
11. What problems do you foresee in selection of tree heights in GP? How does Koza’s rampedhalf-and-half method address this problem?
10. What is the major distinction between GA and GP?
9. A genetic operator not discussed for GA is inversion. Choose two sites randomly on a chromosome: 10^0100^11, and then reverse all characters between these two points as:10001012.a. First, what
8. Suppose that you are using a GA to solve an instance of the TSP. What precautions must be taken when performing crossover?
7. One selection algorithm not discussed is that of miser selection, in which the worst member of a population is selected to participate in reproduction. What advantages do you foresee for this
6. Which operator do you believe is more useful to a GA—crossover or mutation? Defend your assertion.
5. Explain the genetic operators: selection, crossover, and mutation used in GA.
4. How does the temperature parameter T help SA to balance exploitation and exploration?
3. What is the disadvantage of favoring exploitation over exploration in a search (hint: think of hill climbing)?
2. Define exploitation and exploration in search algorithms.
1. What is the relationship between annealing and SA?
10. Write a program to solve the TSP when n = 10 cities, using a Discrete Hopfield network(consult Exercise #15). As in the previous program, run your program 10 times and discuss your results.
9. Write a program to solve the 4–Queens problem using a Discrete Hopfield Network (consult Exercise #14). Run your program 10 times choosing different units for update on each run.Discuss your
8. Use your program in Programming Exercise 5 to classify irises into three classes—Setosa, Versicolor, and Virginia. [See DVD, Appendix D.2.1]
7. Use your program in Programming Exercise 5 to predict the next week’s gold price based upon the following data. Use the last 25% of the data for validation. [See DVD, Appendix D.2.1]
6. Use your program in Programming Exercise 5 to approximate the function that provides the weight in mg of Wild Australian Rabbits as a function of age (in days). Withhold every third data item in
5. Write a program that can apply the backpropagation algorithm to any two-layer feed forward network.
4. Write a program to implement the backpropagation algorithm to train the two-layer network in Figure 11.38 for the XOR function.
3. Use the Delta Rule to complete the training for the two-input OR function in Example 11.2.
2. There are 24 or 16 Boolean functions of 2 variables. Use the Perceptron Learning Rule to determine how many of these are linearly separable.
1 and ten lie in Class 2 and these classes are linearly separable. Use the Perceptron Learning Rule on this data with a learning rate α = (a) 0.01 (b) 0.1 (c) 0.25 (d) 0.50 (e) 1.0 (f) 5.0.Comment
1. Generate 20 triples of random numbers (60 in total) where each number ∈ [0, 1]. Each triple corresponds to a point in the unit cube. Generate these numbers so that ten triples lie in Class
16. Consider the Hopfield network drawn in Figure 11.67. Calculate the energy for each state and draw the state transition diagram for this network. Identify the stable states (if any).
15. The TSP was also discussed extensively in both Chapters 2 and 3. Specify the architecture of a Discrete Hopfield Network to solve a small instance of this problem, say n = 4 cities.Hint—use an
14. The n-Queens Problem was discussed extensively in Chapter 2. Specify the architecture of a Discrete Hopfield Network to solve this problem when n = 4.
13. The n-Rooks Problem is to place n rooks on an n × n chessboard so that these pieces are nonattacking. A rook can attack any piece on the same row or column. A solution to the 4–rooks problem
12.a. Propose a design for a BPN to help make health insurance premium decisions for an insurance company. Prospective clients for health insurance are to be classified as lowrisk or high-risk
11.a. Solve Exercise 6 using the Delta Rule. Train the neuron for one epoch.b. What is the stopping criterion for this learning rule?
10. Which of the following sets of points are linearly separable?a. Class 1: {(0.5, 0.5, 0.5), (1.5, 1.5, 1.5)}Class 2: {(2.5, 2.5, 2.5), (2.5, 2.5, 2.5)}b. Class 1: {(1, 1, 0), (2, 3, 1), (3, 2,
9. Suppose that you have a single TLU with n inputs; in other words, unaugmented input vector xi , will have n components, and you are performing Perceptron Learning. How many epochs ust you wait
8. A TLU is being trained using the Perceptron Learning Rule. Input vector and weight vector appear as in Figure 11.65.The target t for the current pattern = 1, however, the actual output of the unit
7. Use the Perceptron Learning Rule to learn the majority function on three inputs where the second input x2 is held fixed at 1. Maj(x1,x2,x3) = 1 whenever two or three of x1, x2, and x3 are equal to
6.a. Use the Perceptron Learning Rule to train a neuron to learn the two-input function depicted in Figure 11.64. Use an augmented input vector. Let the initial weight values be w1 = 0.1, w2 = 0.4
5. Prove that the two-input XNOR function cannot be implemented with a single perceptron. Your proof should use a system of inequalities.
4. It is a well-known physiological phenomenon that if a cold stimulus is applied to a person’s skin for a very short period of time, the person will perceive heat. However, if the same stimulus is
3. What function F is computed by the McCulloch-Pitts network in Figure 11.62?
2. Design a McCulloch-Pitts network for the three-input minority function where Min(x1,x2,x3) equals 1 whenever only one or none of the inputs equal 1, in other words, Min(x1, x2, x3) =
1. Draw a McCulloch-Pitts network to implement the sum function S for a full adder where S(ABCi) = A′B′Ci + A′BCi′ + AB′Ci′ + ABCi.
13. Contrast supervised and unsupervised learning.
12. What are the differences between biological and artificial neurons (neural networks) in terms of both structure and functionality?
11. The human brain consists of between 10 billion (1010) and 100 billion (1011) neurons. Once we understand the workings of the human brain 1 and we construct full-scale software and/or hardware
10. What is the difference between offline and batch training?
9. Why will both the Delta Rule and backpropagation continue to have some error, whereas the Perceptron Learning Rule halts when there is no error present?
8. The backpropagation algorithm is often referred to as the generalized Delta Rule. Why do you think this is so?
7. What information does the dot product of x with w provide in an ANN? How is this information used in the following:a. The Perceptron Learning Rule?b. The Delta Rule?c. Backpropagation?
6. The learning rate is a constant between 0 and 1, in other words, 0 < α ≤ 1. Since a larger learning rate results in faster learning, why not use large values for α?
5. A single-layer neural network cannot implement a function that is not linearly separable. Is this a serious drawback? Explain.
4. Why might stress in humans be considered a nonlinear phenomenon?
3. Nonlinear systems do not obey the proportionality relation between input and output changes as do linear systems. Consider an artificial neuron with a threshold θ = 0.50. Argue that this neuron
2. In a linear system, the output is proportionally related to the input, in other words, small changes in the input produce correspondingly small changes in output and similarly for large input
1. In Section 1 we portray an ANN as a black box. What limitations does this opaqueness impose on their utility?Next Page
6. If your decision tree does not converge, what is needed here?i. More input data?ii. Attributes that do a better job of separating hypotheses?
5. Refer first to exercise 3. Design a decision tree to distinguish between bronchitis, pneumonia, and TB.
4. Table 10.2 contains data about two medical conditions:A person has a cold vs. a person has the flu. Use ID3 to build a decision tree to determine which affliction a person suffers from based upon
3. Test your trees from programming exercises 1 and 2 with:Spaghetti carbonara where sauce = white, contains meat = true, and contains seafood is false.What result did you obtain? Is it what you
2. Add some noisy data to the tree construction program above.Comment on what happens. For example:Sauce = red, contains meat = true, and contains seafood = false, but likes = no.Penne with Bolognese
1. Use ID3 to confirm the final form of the decision tree for pasta preferences in Section 10.7.
3. Entropy is also defined when classification is into three or more classes.The entropy of a set S relative to n distinct classes is defined as:Entropy (S) = 2 1 log n i i i p p = Σ−where pi is
2. Calculate the entropy for each of the following sets:a. [6(+), 11(-)]b. [1(+), 9(-)]c. [2(+), 12(-)]
1. Design a decision tree for the following Boolean functions:a. a ˅ (b ˄ ~ c)b. majority (x,y,z)
13. Are decision trees a lazy or eager learner? Explain your answer.
12. Suggest a possible method to handle attributes with continuous values.
11. Why does choosing that attribute with the largest information gain favor construction of shorter decision trees?
10. When calculating the entropy of a set, why are logarithms calculated with a base equal to 2?
9. Look on the web and find several other areas where decision trees are used.
8. Give an example of where you have used Occam’s Razor in your everyday experience.
7.a. What is Occam’s Razor?b. Does it claim that the shortest hypothesis is always the best?
6. When performing curve fitting, why is a function that passes through all the points in a training set not necessarily the best hypothesis?
5. Describe inductive learning.
4. Why is feedback important to an agent?
3. Describe three different forms of feedback in ML systems.
2. List several ML paradigms.
1. What is machine learning and why is it such an important subfield of AI?
15. Research expert systems built in the past 5 years. What are their features?How are they different from earlier expert systems described in this chapter?
14. How would expert systems need to perform to pass the Turing Test?
13. One of the criticisms of expert systems has been that they are conducive to the creation of microworlds (e.g. by Professor Hubert Dreyfus, see Chapter 6, Section 6.8, p.185). Explain why you do
12. Who owns expert systems? Expert systems have long been considered a major success story from the field of artificial intelligence; however, they have also become somewhat standard and common.
11. Explain why MYCIN was such an important program to all future expert systems and shells.

Showing 3300 - 3400 of 4588

First
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Last

Services

Sitemap
Fun
Definitions
Become Tutor
Study Help Categories
Recent Questions
Expert Questions

Company Info

Security
Copyrights
Privacy Policy
Terms & Conditions
SolutionInn Fee
Scholarship

Get In Touch

About Us
Contact Us
Career
Jobs
FAQ
Campus Ambassador

Secure Payment

Download Our App

© 2024 SolutionInn. All Rights Reserved