0

Suppose we have a DDPG algorithm. The actor has N input nodes, two hidden layers with J nodes, and S output nodes. The critic has N+S input nodes, two hidden layers with C nodes, and one output node. How does the time complexity of this algorithm could be calculated??

farnad
  • 1
  • Hello. Welcome to Artificial Intelligence Stack Exchange. My question to you is: what have you tried so far to answer your own question? Do you know what the time complexity is? – nbro Nov 24 '21 at 10:53
  • Thanks. yes, I know what it is, and I've found some ways to calculate the neural network's time complexity. But, I wonder is there a difference between neural network complexity and Rl algorithms? also in a paper related to my work there was a Q-learning time complexity analysis. But I couldn't find anything for actor-critic methods like DDPG. – farnad Nov 24 '21 at 14:01
  • I would recommend that you first provide the pseudocode of the algorithm (and the link to the paper that contains the algorithm/pseudocode) that you referring to, so that to make sure we're talking about the same thing. – nbro Nov 24 '21 at 14:05
  • I thought my question was clear enough. It's the DDPG paper [link](https://arxiv.org/abs/1509.02971). and I've found this very helpful for neural networks [link] (https://ai.stackexchange.com/questions/5728/what-is-the-time-complexity-for-training-a-neural-network-using-back-propagation) – farnad Nov 24 '21 at 14:10
  • So, which algorithm in that paper are you talking about? Algorithm 1? As I said above, please, edit your post to directly include these details there. – nbro Nov 24 '21 at 15:06

0 Answers0