mark mark
  • Member for 1 year, 7 months
  • Last seen this week
Does DQN generalise to unseen states in the case of discrete state-spaces?
Accepted answer
1 votes

An environment is said to have a discrete state-space, when the number of all possible states of the environment is finite. For example, $3\times3$ Tic-tac-toe game has a discrete state-space, since ...

View answer
Does this tutorial use normalization the right way?
0 votes

You could rank every pixel in terms of brightness before normalization and after to verify that the ranks are preserved. Correct normalization would preserve the ranks. From the pictures 1 and 2, it ...

View answer
Loss randomly changing, incorrect output (even for low loss) when trying to overfit on a single set of input and output
Accepted answer
0 votes

Softmax activation always adds up to 1, because it's designed to deal with probabilities (in problems of classification, those probabilities represent how likely the network thinks an object belongs ...

View answer