April 2019
Intermediate to advanced
212 pages
5h 34m
English
In this section, we'll go through how to update the Q-table by calculating the Q-value of the state we're in and the action we're taking.
As a learning agent, when we're in a state and decide to take an action, our job is to choose the optimal next action based on what we know about our situation so far. When we're at a corner and want to decide whether to turn right or left, we want to look back to the last time we were at that corner, and the time before that, and what happened when we made each decision.
Which decision yielded the highest rewards for us over time?
The Q-table is our agent's way of storing this information and calling it up again when it's needed. It is ...
Read now
Unlock full access