This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: CS 188 Spring 2010 Introduction to Artificial Intelligence Final Exam INSTRUCTIONS • You have 3 hours. • The exam is closed book, closed notes except a twopage crib sheet. • Please use nonprogrammable calculators only. • Mark your answers ON THE EXAM ITSELF. If you are not sure of your answer you may wish to provide a brief explanation. All short answer sections can be successfully answered in a few sentences AT MOST. First name Last name SID Login For staff use only: Q1. Search /14 Q2. Naive Bayes /15 Q3. Hidden Markov Models /15 Q4. Machine Learning /13 Q5. Markov Decision Processes /10 Q6. Short answer /33 Total /100 1 Q1. [14 pts] Search For the following questions, please choose the best answer (only one answer per question). Assume a finite search space. (a) [2 pts] Depthfirst search can be made to return the same solution as breadthfirst search using: (i) Iterative Deepening (ii) A closed list/list of nodes that have been expanded (iii) A heuristic function (iv) This is not possible (b) [2 pts] A * search can be made to perform a breadthfirst search by setting (fill in correct values): 1. for all nodes, heuristic = 2. for all nodes, edgecost = (c) [2 pts] You run A * search using a heuristic function which you know to be admissible and consistent. Your friend claims he has a search algorithm that is guaranteed to not expand more nodes than your algorithm (and in fact often expands far fewer in practice). He also tells you that his algorithm is guaranteed to find the optimal path. Could the algorithm your friend claims to have exist? (circle one): yes no Explain: (d) [2 pts] Depth first search using a closed list/list of nodes that have been expanded is: (i) Optimal (will find a shortest path to goal) (ii) Complete (will find a path to goal if at least one exists) (iii) Both optimal and complete (iv) Neither optimal nor complete 2 Consider a grid, a portion of which is shown below: You would like to search for paths in this grid. Unlike in Pacman, it is possible to move diagonally as well as horizontally and vertically. The distance between neighboring grid squares (horizontally or vertically) is 1, and the distance between diagonally adjacent grid squares is √ 2. (e) [2 pts] Is the euclidean distance an admissible heuristic? The euclidean distance between two points ( x 1 ,y 1 ) and ( x 2 ,y 2 ) is p ( x 2 x 1 ) 2 + ( y 2 y 1 ) 2 . (f) [2 pts] The Manhattan distance is not an admissible heuristic. Can it be made admissible by adding weights to the x and y terms? The Manhattan distance between two points ( x 1 ,y 1 ) and ( x 2 ,y 2 ) is  x 2 x 1  +  y 2 y 1  . A weighted version with weights α and β would be α  x 2 x 1  + β  y 2 y 1  . Specify the (possibly empty) set of pairs of weights ( α,β ) such that the weighted Manhattan distance is admissible....
View
Full Document
 Spring '08
 Staff
 Artificial Intelligence, Manhattan, A* search algorithm, Viterbi algorithm, Consistent heuristic, Hidden Markov model

Click to edit the document details