State spaces
When we are given information about the state of a system, such as a multi-armed bandit environment, the state space becomes part of our knowledge of the problem, and it allows us to adjust our probability expectations of the outcomes we can expect from taking different actions from a particular state.
For example, if we have preliminary information about a potential customer, such as their gender or age, we can take that information into account by incorporating it into the state space when deciding what ads to show them.
In other words, the gender and age of a customer become state variables. As we observe the user behavior of an individual person, we can adjust the state information we have about them to reflect these observations. ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access