CSC 261 : Exam 1

CSC 261 - Artificial Intelligence - Weinman

1  Learning Outcomes

• Define a rational agent
• Determine whether an agent's behavior is rational in a given environment
• Distinguish and characterize the four agent architectures
• Justify an appropriate agent program for a given task environment
• For a a given task environment,
  • give a a PEAS description, and
  • categorize the properties (e.g. episodic, deterministic, etc.)
• Conduct a general tree search and graph search on a given state space
• For each specific tree search (depth-first, breadth-first, uniform-cost, greedy, A*),
  • distinguish/identify the frontier sorting criterion,
  • conduct the specific search for a given state space, and
  • identify the frontier enqueueing function,
• Justify the completeness, optimality, time complexity, and space complexity of breadth-first, depth-first, uniform-cost, depth-limited, and iterative depeening searches
• Identify and distinguish the three strategies for heuristic design
• Justify an application of a particular heuristic design strategy to a given problem
• Distinguish between hill-climbing search variants, simulated annealing, and local beam search
• Understand the role of the gradient and step size in local search of continuous spaces
• Create a contingency plan by generating an AND-OR search tree for a non-deterministic environment
• Generate a state space graph of belief states for a partially observable environment or sensorless agent
• Run PREDICT and UPDATE steps for partially observable, non-deterministic environments
• Give the sequence of states visited and/or actions chosen for an agent running the ONLINE-DFS-AGENT function.
• Determine the node values in a Minimax search tree
• Identify the branches pruned by alpha-beta search

2  Study Problems

Chapter 2

1-6, 10-13

Chapter 3

2, 3, 6, 8, 9(a,c), 11, 12, 14, 15, 18, 21, 25, 26, 29

Chapter 4

1, 5, 7, 8, 9, 12, 14

Chapter 5

1, 3, 7

Chapter 2

2

Chapter 4

1, 2, 3, 12