a$)$ How are rewards and returns connected in Deep Re$-$inforcement Learning?

b$)$ Consider the RL agent learning the game of tic$-$tac$-$toe, by playing against different randomly chosen components. Consider the temporal difference rule being used in this context. can alpha be used to encourage exploration? why or why not?

c$)$ Write the model$-$ based and model free algorithm for reinforcement learning.

$V(s_t)larrV(s_t)+[V(s_{t+1})-V(s_t)]$