0 able to see clearly only objects that are not close
1 able to make wise judgments about the results far in the future of an action you take now:
a farsighted proposal
The agent is said to be farsighted since it looks far ahead of time when evaluating a state-action pair.
For such a reward function, a "farsighted" approach for the return computation has to be adopted.
However, central banks may not be that farsighted; they may simply postulate a certain amount of inertia on how private agents forecast.
The learning agent was farsighted as in the previous implementation.
For the simple reward function, a farsighted learning agent was used.
In the third analysis, we modify the reinforcement learning algorithm so that it only uses a myopic learning agent instead of a farsighted one.
In summary, this implementation produces better result than the farsighted one.
At the other extreme, when = 0, the fishers will treat the far-distant future as being equally valuable as today; that is, they will be very farsighted.