Consider a 100x100 grid world. The agent's task is to get a goal object. The object can move to any of its 4 neighboring locations ( up, down, left, right). The agent has a sensor that it can use to perceive the brightness at the current locations. The location with the goal object has a brightness value of 15. Locations that are less than 5 steps away will have a brightness value of 10. Locations that are 6 to 10 steps away will have a brightness value of 5. all other locations will have a brightness value of 0.
A) Describe the environment (Observable? Deterministic?Episodic?Static?Discrete?Single-agent?)
B) Which of these heuristics are admissible? Which are not? If both admissible which heuristics dominates the other? Justify the answer.
--> h1(n): The straight-line distance between n and goal
--> h2(n): The Manhattan distance between n and goal
C) Define a maximally admissible heuristics that use the brightness values observed by the sensors to estimate distance.
h1(n) = sq root of [(xg-xn)^2 + (yg-yn)^2]
h2(n) = |xg-xn| + |yg-yn|
Bu iş için 2 freelancer ortalamada $28 teklif veriyor
Just saw this while browsing. I have the answer ready. In fact I created an account only so that I could respond.