Markov Decision Process

about | blog | config | notes | github

Fully Observable Environments are Environments where the Agent is able to fully see the internal environment state (everything). Formally,

\[O_t = S^a_t = S^e_t\]

As every state is a Markov State, a Fully Observable Environment is a Markov Decision Process.

Created: 2022-02-03

Emacs 26.1 (Org mode 9.5)