[UPDATE] : Unit 3 DQN

[pdx97]

What do you want to improve?

• You could add a bonus unit in Unit Three where you talk about the Q Learning and its importance as the discount factor determines how much the reinforcement learning agents cares about rewards in the distant future relative to those in the immediate future.

• If γ=0, the agent will be completely myopic and only learn about actions that produce an immediate reward.

• If γ=1, the agent will evaluate each of its actions based on the sum total of all of its future rewards

• Explain the typo/error or the part of the course you want to improve Add Gamma and its significance . I believe its a important thing to tell as DQNs are very sensitive to Hyperparameters.

[simoninithomas]

Hey there 👋 , we already mentioned this on Q-Learning, which is required to study DQN, you think we should also add it to the DQN unit too?

[pdx97]

ohh then its fine , if you have already mentioned it in the Q learning , I might have missed it . 😊

[simoninithomas]

Ok, in that case I'm closing the issue 🤗