Q5. Feature-based Representation (20 points) Consider the following feature based representation of the Q-function: Q(8,0) = wifi(sa) + w2f2(8.a) with: fi(s, a) = 1/ (Manhattan distance to nearest dot after having executed action a in state s) $2(, a) =(Manhattan distance to nearest ghost after having executed action a in state s) Q5.1. Assume wi = 1 and W2 = 10. Assume that the red and blue ghosts are both sitting on top of a dot. Provide the values of Q(s, west) and Q(s, south). Based on this approximate Q-function, which action would be chosen? Q5.2. Assume Pac-Man moves West. This results in the state s' shown below. Pac-Man receives reward 9 (10 for eating a dot and -1 living penalty). Provide the values of Q(s', west) and Q(s', east). What is the sample value (assuming 7 = 1)? Q5.3. Now provide the update to the weights. Let a=0.5. Q5. Feature-based Representation (20 points) Consider the following feature based representation of the Q-function: Q(8,0) = wifi(sa) + w2f2(8.a) with: fi(s, a) = 1/ (Manhattan distance to nearest dot after having executed action a in state s) $2(, a) =(Manhattan distance to nearest ghost after having executed action a in state s) Q5.1. Assume wi = 1 and W2 = 10. Assume that the red and blue ghosts are both sitting on top of a dot. Provide the values of Q(s, west) and Q(s, south). Based on this approximate Q-function, which action would be chosen? Q5.2. Assume Pac-Man moves West. This results in the state s' shown below. Pac-Man receives reward 9 (10 for eating a dot and -1 living penalty). Provide the values of Q(s', west) and Q(s', east). What is the sample value (assuming 7 = 1)? Q5.3. Now provide the update to the weights. Let a=0.5