HW1_Solutions

# 3 exercise 4 part i a performance measure the

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Performance Measure – The percentage of wins over the long run (maximize) Environment – All the cards in the game (the hands, the deck, etc.) and the dealer Actuators – choose whether to “hit” or “stand” Sensors – See the cards in the agent’s hand b) A simple reflex agent would need to choose whichever of “hit” or “stand” has the best chance of winning based on all the cards played so far. This would be impossible for a simple- reflex agent to do because it cannot remember the cards played in previous rounds. c) Condition (Percept) Action CalcSum(agent) &lt;= 17 Hit CalcSum(agent) &gt; 17 Stand d) Using seed 0, simulating 50 rounds: Agent Type Percentage of Wins Simple reflex mimic dealer 36% Part II a) Yes....
View Full Document

Ask a homework question - tutors are online