CS 580 Spring 2006 Dr. Duric

Final Review 1. Consider the following game tree. MAX

A

B

D

C

E

3

F

G

H

5

4 I

K

J 5

M 0

7

L 8

N 7

(a) Find the best move for the MAX player using the minimax procedure. (b) Perform a left-to-right alpha-beta pruning on the tree. Indicate where the cutoffs occur. (c) Perform a right-to-left alpha-beta pruning on the tree. Discuss why different pruning occurs. 2. For each of the following pairs of atomic sentences, give the most general unifier, if it exists. a. P(A,B,B), P(x,y,z) b. Q(y,G(A,B)), Q(G(x,x),y) c. Older(Father(y),y), Older(Father(x),John) d. Knows(Father(y),y), Knows(x,x) e. Ancestor(x,y), Ancestor(John,Father(John)) 3. The function cons(x,y) denotes list formed by inserting the element x at the head of the list y. We denote the empty list by N il; the list (2) by cons(2, N il); the list (1, 2) by cons(1, cons(2, N il)); and so on. The formula M ember(e, l) is intended to mean that e is a member of the list l. We have the following axioms: – (∀x, y)[M ember(x, cons(x, y))] – (∀x, y, z)[M ember(x, y) ⊃ M ember(x, cons(z, y))]

Prove M ember(b, cons(a, cons(b, nil))) from these axioms by the method of resolution refutation. 4. Dr. Somebody, Dr. Anybody, and Dr. Nobody are computer scientists. We know the following facts about them: 1. Dr. Somebody is an associate professor. 2. Dr. Nobody is an assistant professor and has published papers with Dr. Anybody. 3. Dr. Anybody is either an associate or an assistant professor (but not both) and has published papers with Dr. Somebody. Use resolution refutation to prove that an assistant professor has published papers with an associate professor; that is, prove (∃x, y)[Assistant(x) ∧ Associate(y) ∧ P P W (x, y)]. 5. We are given the following paragraph: Tony, Mike, and John belong to the Alpine Club. Every member of the Alpine Club is either a skier or a mountain climber or both. No mountain climber likes rain, and all skiers like snow. Mike dislikes whatever Tony likes and likes whatever Tony dislikes. Tony dislikes rain and snow. Represent this information by predicate-calculus sentences in such a way that you can represent the question “Is there a member of the alpine club who is a skier but not a mountain climber?” as a predicate-calculus expression. Use resolution refutation to answer it. 6. A simple version of the nim game is played as follows: Two players alternate in removing stones from three piles initially containing one, two, and three stones, respectively. The player who picks up the last stone wins. Each player can pick one or more stones from a single pile; at least one stone has to be picked every time. At every turn the players can pick from different piles. Show, by drawing the game tree, which player can always win. Is it necessary to generate the whole tree to find a winning strategy? 7. Consider the following statements: (1) Whoever can read is literate (∀x[R(x) ⇒ L(x)]). (2) Horses are not literate (∀x[H(x) ⇒ ¬L(x)]). (3) Some horses are intelligent (∃x[H(x) ∧ I(x)]). Use these statements to prove the following statement by resolution: (4) Some who are intelligent cannot read (∃x[I(x) ∧ ¬R(x)]). 8. You are given the following table:

Outlook sunny sunny sunny sunny sunny overcast overcast overcast overcast rain rain rain rain

Temperature warm hot hot warm warm warm hot cool hot warm cool warm cool

Humidity (%) < 75 > 75 > 75 > 75 < 75 > 75 > 75 < 75 < 75 > 75 < 75 < 75 > 75

Windy? yes yes no no no yes no yes no yes yes no no

Class Play Don’t Play Don’t Play Don’t Play Play Play Play Play Play Don’t Play Don’t Play Play Play

(a) Build a decision tree to distinguish the tow classes. Use the information theory to decide which attribute should be used at the root of the decision tree (10p for this step). (b) Indicate the ’Play’ concept represented by the tree. 9. We represent the statement that everything is representable in the predicate calculus as ∀x, represents(pc, x). General Problem Solver (GPS) is a system for automated problem solving. We represent the statement that all problems representable in predicate calculus are solvable using GPS as ∀x, (problem(x) ∧ represents(pc, x)) ⇒ solves(gps, x). Now using these two statements and the fact that the Traveling Salesperson Problem (TSP) is a problem (problem(tsp)), prove that GPS solves it. 10. Attempt to unify the following pairs of expressions. Either show their most general unifiers or explain why they will not unify. (a) p(X,Y) and p(a,Z) (b) p(X,X) and p(a,b) (c) ancestor(X,Y) and ancestor(bill,father(bill)) (d) ancestor(X,father(X)) and ancestor(david,george) 11. Consider the sentence “Heads I win; tail you lose.” We can represent this sentence plus associated domain knowledge in the propositional logic using the following proper axioms, where Heads, Tails, WinMe, and LoseYou are propositional variables: Heads ⇒ W inM e T ail ⇒ LoseY ou ¬Heads ⇒ T ails LoseY ou ⇒ W inM e a) Determine if it is possible to prove WinMe using the rule of inference modus ponens and these four axioms. b) Convert each of the four axioms to a disjunction of literals. c) For each of the resulting disjunctions, specify if it is a Horn clause.

d) Determine if it is possible to prove WinMe using the resolution rule of inference and the four axioms written as disjunctions of literals. 12. We have defined four different binary logical connectives (∧, ∨, ⇒, ⇔). a. Are there any others that might be useful? b. How many binary connnectives can there be? c. Why are some of them not very useful? 13. Consider the following problem: The rules in the National Zoo forbid visitors to feed animals. A prairie dog, an animal, has some candy, and all of its candy was given to it by BB, who is a Zoo visitor. a. Represent these facts as first-order logic clauses. b. Use backward-chaining to prove that BB broke Zoo rules. c. Use forward-chaining to prove that BB broke Zoo rules. 14. Build a decision tree to distinguish two classes (’reads’ and ’skips’) from the following table: Example e1 e2 e3 e4 e5 e6 e7 e8

Action skips skips skips reads reads reads skips skips

Author known known unknown known unknown known unknown unknown

Thread new follow up follow up follow up new new new follow up

Length long long long short short short long short

Use the information theory to decide which attribute to use at each non-terminal node of the tree. 15. Sam, Clyde, and Oscar are rabbits. We know the following facts about them: 1. Sam is pink. 2. Clyde is gray and likes Oscar. 3. Oscar is either pink or gray (but not both) and likes Sam. Use resolution refutation to prove that a gray rabbit likes a pink rabbit; that is, prove (∃x, y)[Gray(x) ∧ P ink(y) ∧ Likes(x, y)]. 16. The function cons(x,y) denotes list formed by inserting the element x at the head of the list y. We denote the empty list by nil; the list (2) by cons(2, nil); the list (1, 2) by cons(1, cons(2, nil)); and so on. The formula Last(l, e) is intended to mean that e is the last element of the list l. We have the following axioms: – (∀u)[Last(cons(u, nil), u)] – (∀x, y, z)[Last(y, z) ⊃ Last(cons(x, y), z)]

1. Prove the following theorem from these axioms by the method of resolution refutation: (∃v)[Last(cons(2, cons(1, nil)), v)] 2. Use answer extraction to find v, the last element of the list (2, 1). 17. The set of inputs x1 , x2 define a 2-dimensional space. A boolean function defining a concept is said to be linearly separable if there exists a hyperplane (line) dividing the space into inputs for which the function produces a 1 and inputs for which the function produces a 0. a) Which of these four boolean functions are linearly separable? or and xor equal Justify your answer. b) What needs to be done to make concepts which are not linearly separable learnable by the perceptron learning rule. Illustrate your answer using a boolean function that is not linearly separable. 18. A simple version of a “Kayles game” is played as follows: Two players have in front of them a single contiguous sequence of objects, say 5 pennies, which are placed next to each other so that the first penny touches the second, the second touches the third, the third touches the fourth, and the fourth touches the fifth penny. The first player removes 1 or 2 pennies whose sides that are touching (for example, the first player can remove any single penny, or pennies 1 and 2, or pennies 2 and 3, or pennies 3 and 4, or pennies 4 and 5). Each player alternatively thereafter removes 1 penny or 2 pennies whose sides are touching. (Note: If the first player removes penny 2, the pennies 1 and 3 will have a gap between them and cannot be removed at the same time.) The last player to pick a penny (or two) loses. Show, by drawing a game tree, whether any of the players can always win. You can ignore symmetries and obvious losing plays if there is a winning play available, such as picking both pennies when there are 2 pennies left. 19. Consider a vocabulary with only four propositions, A, B, C, and D. How many models are there for the following sentences? a. (A ∧ B) ∨ (B ∧ C) b. A ∨ B c. A ⇔ B ⇔ C Hint: In each sentence all four propositions have to be considered. 20. Build a decision tree to distinguish two classes (’+’ and ’–’) from the following table: food herbivore carnivore omnivorous herbivore omnivorous carnivore carnivore omnivorous

medium land land water amphibious air land land land

type harmless harmful harmless harmless harmless harmful harmless moody

class mammal mammal fish amphibian bird reptile reptile mammal

+ – + – – + – +

description deer (e1) lion (c1) goldfish (e2) frog (c2) parrot (c3) cobra (e3) lizard (c4) bear (e4)

Use information theory to decide which attribute should be used at each nonleaf node of your decision tree. 21. A simple version of the nim game is played as follows: Two players alternate in removing stones from two piles initially containing several stones each. The player who picks up the last stone wins. At any given turn a player can pick one or more stones from a single pile; at least one stone has to be picked every time. a) Show, by drawing a game tree, which player can always win if the piles have two and two stones. b) Show, by drawing a game tree, which player can always win if the piles have two and three stones. You can use your result from a). c) What can you say about a game in which the piles have m > 0 and n > 0 stones? 22. You are building a decision tree which tells us whether we should go to a restaurant based on various attributes: how many patrons it has, whether or not the food is cheap, and what type of food is served. The information is: Example 1 2 3 4 5 6

patrons some empty some some full empty

cheap yes yes no yes no yes

type french thai burger thai french italian

go yes no yes yes no yes

a) Show the tree after asking attribute questions in this order: type?, cheap?, patrons? b) Express the concept represented by the decision tree as a first order logical expression. c) What is the first attribute question you should ask and why? Show your work. 23. A simple version of the nim game is played as follows: Two players alternate in removing stones from three piles initially containing two stones each. The player who picks up the last stone wins. Each player can pick one or more stones from a single pile; at least one stone has to be picked every time. At any turn the player can pick from any one pile. Show, by drawing the game tree, whether a player can always win, and if so which one. Is it necessary to generate the whole tree to find a winning strategy? 24. A typical belief network with conditional probabilities is given in the following figure:

P(C) .50 Cloudy

C P(S|C) T .10 F .50

C P(R|C) T .80 F .20

Rain

Sprinkler

Wet Grass

S R P(W|S,R) T T F F

T F T F

.99 .90 .90 .01

The letters C,R,S and W stand for Cloudy, Rain, Sprinkler, and Wet Grass, respectively. All variables (nodes) are Boolean, so the probability of, say, ¬A in any row of its table is 1−P (A). a) Assuming that the nodes are introduced in the following order Wet Grass, Sprinkler, Rain and Cloudy construct a corresponding belief network. Show which probabilities need to be specified. b) Compute probabilities P(W ) and P(S|W ). 25. A typical belief network with conditional probabilities is given in the following figure:

Burglary

P(B)

.001

Alarm

JohnCalls

P(E)

Earthquake

A

P(J)

T F

.90 .05

B

E

P(A)

T T F F

T F T F

.95 .94 .29 .001

MaryCalls

.002

A

P(M)

T F

.70 .01

The letters B,E,A,J, and M stand for Burglary, Earthquake, Alarm, JohnCalls, and MaryCalls, respectively. All variables (nodes) are Boolean, so the probability of, say, ¬A in any row of its table is 1 − P (A).

a) Assuming that the nodes are introduced in the following order JohnCalls, MaryCalls, Alarm, Earthquake, and Burglary construct a corresponding belief network. Show which probabilities need to be specified. b) Compute probabilities P(A), P(E|A), and P(B|A, E).

Final Review - GMU CS Department

associated domain knowledge in the propositional logic using the following .... on various attributes: how many patrons it has, whether or not the food is cheap, ...

94KB Sizes 5 Downloads 316 Views

Recommend Documents

Manufacturing Compromise: The Emergence of ... - GMU CS Department
exploit kits themselves are not new, dating back to at least MPack in 2006 [34], there is little .... states that they buy any type of traffic from affiliates, regard- less of quality ... Traffbiz: Traffbiz is a Russian Traffic-PPI service. They also

Department of Psychiatry-Final Report.pdf
embraced the change and improvements proposed and have gone above and beyond what has been. requested of them. They have embraced a new culture of ...

CHEMISTRY DEPARTMENT time table final listupdated.pdf ...
Biochemistry Third. Quantum Chemistry Fourth. Page 1 of 1. CHEMISTRY DEPARTMENT time table final listupdated.pdf. CHEMISTRY DEPARTMENT time table ...

84-cs final june 16 DIRECT TAX amendments.pdf
applied. Surcharge = 12% , If the total income exceeds ` 1crores. Education Cess: 2% on Total Tax. Secondary Higher Education Cess: 1% on Total Tax.

M.Sc(CS)_Semester System(Only for University Department) Ses.pdf
Page 3 of 23. M.Sc(CS)_Semester System(Only for University Department) Ses.pdf. M.Sc(CS)_Semester System(Only for University Department) Ses.pdf. Open.

Final Review Energy.pdf
What is Total Mechanical Energy equal to? 7. What two factors does the amount of gravitational potential energy depend on? (look at the formula). 8. What two ...

Final GRADUATE PROGRAM PRIORITIZATION Review Tiers.pdf ...
Final GRADUATE PROGRAM PRIORITIZATION Review Tiers.pdf. Final GRADUATE PROGRAM PRIORITIZATION Review Tiers.pdf. Open. Extract. Open with.

Math 114-Review Final Exam
The equation that models its height, h feet, off the ground t seconds after it was fired is h = - 16t2 + 40t. (a) How high is the rocket 1.5 seconds after it was fired?

C1730 Final Review Questions.pdf
Sign in. Loading… Whoops! There was a problem loading more pages. Retrying... Whoops! There was a problem previewing this document. Retrying.

INTRODUCTION TO ROBOTICS final exam REVIEW SHEET.pdf ...
... Data. Programming to accumulate and react through the use of variables. Page 1 of 1. INTRODUCTION TO ROBOTICS final exam REVIEW SHEET.pdf.

SPRING FINAL REVIEW key.pdf
Download. Connect more apps... Try one of the apps below to open or edit this item. SPRING FINAL REVIEW key.pdf. SPRING FINAL REVIEW key.pdf. Open.

Pre-16 Final Review Problems
2) When some data is plotted on semi-log (log-linear) paper, the line logy = log 3− xlog4 resulted. What is the functional relationship between x and y (solve for y as a function of x with no logs!) 3) Graph. Include intercepts and asymptotes: a) W

Final Review-2072.pdf
wf/f !* cf}lrTo 5}g\, olb /fVg} kg{] ePdf æpkwf/f -@_Æ clgjfo{ ?kdf x6fpg' kg{] . wf/f !( M df æblIf0f Pl;ofnL If]lqo ;xof]u ;+u7g -;fs{_ If]qeGbf aflx/Æ jfSof+z x6fpg'.

Final Review 2016.pdf
Directions: Use the 2nd Derivative Test to find the local extrema for the given function. 76. h(x) = x. x. 2+2. 77. f(x) = x. 3 − 3x. 2 − 9x + 1 on. 78. k(x) = √x. 2 − 4. 3. 79. b(x) = ln x. x. 2. 80. v(x) = x + 3x. 2. 3. 81. m(x) = x + e. âˆ

Final Exam Review Period
(7.62 x 107) – (8.9 x 106). 12. Sam fills a water balloon with 7.8 x 10-3 ft3 of water. ... 6m – 4(m +1) = 14. 20. Three tables are given below. Decide if the table ...

Infinite Algebra 2 - FINAL Review
13) f (t)= -2t - 3 g(t)= t2 - 3t. Find (f × g)(t). Change to VERTEX FORM. Sketch the graph of each function. State the DOMAIN and. RANGE. Solve for the ROOTS.

9th PAY REVISION COMMISSION 2010 2 final - Finance Department
Mar 31, 2010 - desirable to continue in the present system or to switch over to ... Should annual increment be an automatic ... Any alternative system to suggest, specify. 17. .... Email. : Tele phone Number. : www.finance.kerala.gov.in /prc9.

Introduction to C++ Lab (CS 54) - CS SYLLABI
If you have a documented disability and anticipate needing accommodations in ... request that the Disability Support Services staff send a letter to the instructor ...

CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical ...
CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical Institution.com.pdf. CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical Institution.com.

CS 50 Walkthrough 6
create nodes for them. – put these nodes ... store each le›er i of the word in that node. • fgetc(dptr) is that ... put a pointer to your node that you just malloced there.

Ж Ї - CS, Technion
Some people attribute such capability to ... approach to solve this problem is to pre-program the agent with a domain-specific decision procedure. ... The work presented in this paper takes the pattern-based move-oriented approach similarly.

CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical ...
Complication), Abortion/Miscarriage. Pap Smear (Last pap? ... CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical Institution.com.pdf. CS Blue sheet ...

CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical ...
CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical Institution.com.pdf. CS Blue sheet Mnemonics - USMLE Step 2 CS - www.medical Institution.com.

CS AMRUT PARAKH / 9423040074 ... - CS Amrut K. Parakh
Application for allotment of Director Identification Number ... Application for incorporation of a company ... registered office of the company within the state.