0
$\begingroup$

I know that in any NN architecture, the input data are states, and at the output layer Q-functionality of each action. Tell me please, how to adjust all weights in this case?

$\endgroup$
  • $\begingroup$ Hi. This question seems to be too broad. In fact, I will vote to close it as too broad. However, if you're just looking for the basic idea, then I will provide an answer below. Bear in mind that this question may be closed as too broad anyway. $\endgroup$ – nbro Jan 24 at 12:21
1
$\begingroup$

Have a look at and read the paper Playing Atari with Deep Reinforcement Learning, which describes deep Q-learning (i.e. Q-learning with neural networks). In particular, have a look at algorithm 1 (on page 5). As it is usually the case in deep learning, gradient descent and back-propagation are used to update the parameters (or weights) of the neural network, but they also use another technique, the experience replay buffer, in order to improve the convergence and stability of the training process. In the paper Human-level control through deep reinforcement learning, apart from using the experience replay buffer, they also use a so-called target network to improve stability.

| improve this answer | |
$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.