here in this example we know the terminals but in real world example how do we know the terminals ,caculate Q* and take the optimum actions ?
Hi @Ibrahim_Mustafa Great question.
In a real-world reinforcement learning problem, the terminal states (also known as goal states or absorbing states) are often defined by the problem itself. For example, in a game like chess, a terminal state would be a checkmate or a draw. In a self-driving car scenario, it could be reaching the destination.
Please let me know if this helps!
Ok i understand the first part of my question but how we calculate q* and how to know how to choose actions to perform
Hello @Ibrahim_Mustafa,
I recommend you to go through the rest of the course and finish the assignment first. The lecture will cover how we use neural network to model the Q* for complex problem, and the assignment will apply that neural network technique to learn, as you asked, “how we calculate Q*”, and “how to choose actions to perform”.
Cheers,
Raymond