Deep Reinforcement Learning with Double Q-learning · HackerPapers discussions

Deep Reinforcement Learning with Double Q-learning — mentioned by DevelopingElk

Sun, 15 Jun 2025 05:26:02 GMT

I think the article's analysis of overapproximation bias is correct. The issue is that due to the Max operator in the Q learning noise is amplified over timesteps. Some methods to reduce this bias, such as https://arxiv.org/abs/1509.06461 were successful in improving the RL agents performance. Studies have found that this happens even more for the states that the network hasn't visited many times. An exponential number of states only matters if there is no pattern to them. If there is some structure that the network can learn then it can perform well. This is a strength of deep learning, not…

Paper and discussion history. A mention is not an endorsement.

Deep Reinforcement Learning with Double Q-learning — mentioned by ClintEhrlich

Thu, 24 Sep 2015 15:44:27 GMT

Here is a link to the underlying study, for people who are interested in digging into the data themselves: http://arxiv.org/pdf/1509.06461v1.pdf

Paper and discussion history. A mention is not an endorsement.

Deep Reinforcement Learning with Double Q-learning — HN submission

Wed, 23 Sep 2015 21:49:22 GMT

DeepMind: Deep Reinforcement Learning with Double Q Learning

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.