Passive Reinforcement Learning

CSC 261 - Artificial Intelligence - Weinman



Answer the following questions. Record your answers in your Reading Journal.
  1. Select the sentence from today's reading that you feel highlights the most important consideration for direct utility estimation. Briefly (3-5 sentences) explain your selection.
  2. Select the sentence from today's reading that you feel best distinguishes between direct utility estimation and adaptive dynamic programming. Briefly (3-5 sentences) explain your selection.
  3. How would you explain the TD update Equation (21.3) to a fellow student who knows the perceptron learning rule, Eq. (18.7)?
  4. Which piece(s) of the utility-based agent in Figure 2.14 (p. 54) do you believe are missing from a PASSIVE-TD-AGENT?
  5. Identify (and briefly elaborate on) the sentence, section or concept from the current or previous reading that remains the most confusing to you.