Q1. (Decision Theory) Explain why utility functions are a strictly more expressive preference model than the goals used in deterministic planning.
Answer: Because there is major difference in representation, decisions in stochastic environment are based on beliefs since there is uncertainty present and utility functions are more expressive since they rate how much an agent is happy in different situations, while goals in deterministic planning are “all-or-nothing”.
Q2. (Decision Theory) What is the difference between random variables and decision variables?
Answer: Decision variables represent available actions, while random variables are designated to represent information.
Q3. (Decision Theory) In words, explain how to compute an agent's expected utility for making a single decision.
Answer: To compute agent's expected utility you weigh the utility of the relevant possible worlds by their probability. E(U | D = di ) = Sumw╞ (D = di )(P(w) U(w)) where D is decision and w is world.
Q4. (Decision Theory) How is a sequential decision problem different from a single decision problem?
Answer: In single decision problem agent is making one big decision before acting that may consist of small decisions. In sequential decision problem the agent is making a sequence of decisions by making observations, deciding on the action, executing the action and repeating the process until desired result is not achieved.
Q5. Why does it make sense that in the “single decision” framework, we can still talk about problems (such as the example given in the slides) where there is more than one decision variable?
Answer: Because in “single decision” framework we can still make micro decisions for each of decision variables, and it still will be considered as part one one big decision.