Currently reading and studying Markov Decision Processes.

After seeing search problems as a deterministic way to solve problems, we know that the world is not so certain, and here it comes the stochastic/randomness of the world.

Ideally, a Markov decision process is pretty similar to a search problem, with some differences such as:

After seeing search problems as a deterministic way to solve problems, we know that the world is not so certain, and here it comes the stochastic/randomness of the world.

Ideally, a Markov decision process is pretty similar to a search problem, with some differences such as:

- Set of states
- Set of actions that will help you go to one state to another.
- Transition model, as the probability that given you are in state s and perform action A you end up in state s’
- Policy, is the recommended action in any state
- A reward function, is the reward you obtain after transitioning from one state to another.

There is more to it and I‘ll write more about it as I keep learning.

It’s a fascinating topic and additionally the fundamentals of Reinforcement Learning

It’s a fascinating topic and additionally the fundamentals of Reinforcement Learning