concept reinforcement learning agent in category reinforcement learning

appears as: reinforcement learning agents
Grokking Deep Reinforcement Learning

This is an excerpt from Manning's book Grokking Deep Reinforcement Learning.

Tweetable Tweet

Do some work on your own and share your findings

Every chapter, I’ll give some ideas on how to take what you have learned to the next level. If you’d like, share your results with the rest of the world and make sure to check out what others have done, too. It’s a win-win situation, and hopefully, you take advantage of it.

For this chapter, here are some ideas to explore:

· #gdrl-ch01-tf01: Supervised, unsupervised, and reinforcement learning are es- sential machine-learning branches. And while it is crucial to know the differences, it is equally important to know the similarities. Write a post analyzing how these different approaches compare and how they could be used together to solve the AI problem. All branches are going after the same goal: to create artificial general intelligence, it’s vital for all of us to better understand how to use the tools available.

· #gdrl-ch01-tf02: I wouldn’t be surprised if you don’t have a machine learning or computer science background, yet are still interested in what this book has to offer. One essential contribution is to post resources from other fields that study decision making. Do you have an operations research background? Psychology, philosophy, or neuro-science background? Are you a Control’s folk? Economics? How about you create a list of resources, blog posts, YouTube videos, books, or any other medium and share it with the rest of us also studying decision making?

· #gdrl-ch01-tf03: Some of the text in this chapter has a lot of information that could be better explained through graphics, tables, and other forms. For instance, I talked about the different types of reinforcement learning agents (value-based, poli- cy-based, actor-critic, model-based, gradient-free.) Why don’t you grab some text that is dense, distill the knowledge and share your summary with the world?

· #gdrl-ch01-tf04: In every chapter, I’m using the final hashtag as a ‘catch-all’ hashtag. Feel free to use this one to discuss anything else that you worked on relevant to this chapter. There is no more exciting homework that the one you create for your- self. Make sure to share what you set yourself to investigate and your results.

Write a tweet with your findings, tag me @mimoralea (I’ll retweet,) and use the particular hashtag from the list above to help folks interested find your results. There are no right or wrong results; you share your findings and check others’ findings. Take advantage of this to socialize, contribute, and get yourself out there! We’re waiting for you!

Here is a tweet example:

“Hey, @mimoralea. I created a blog post with a list of resources to study deep reinforcement learning. Check it out at <link>. #gdrl-ch01-tf01”

I’ll make sure to retweet and help others find your work.

Every chapter, I’ll give some ideas on how to take what you have learned to the next level. If you’d like, share your results with the rest of the world and make sure to check out what others have done, too. It’s a win-win situation, and hopefully, you take advantage of it.

6.1        The anatomy of reinforcement learning agents

In this section, I’d like to give you a mental model that most, if not all, reinforcement learning agents fit under. First, every reinforcement learning agent gathers experience samples, either from interacting with the environment or from querying a learned model of an environment. Still, data is generated as the agents learn. Second, every reinforcement learning agent learns to estimate something, perhaps a model of the environment, or possibly a policy, a value function, or just the returns. Third, every reinforcement learning agent attempts to improve a policy, that’s the whole point of RL, after all.