Spinning Up in Deep RL

Plug for the RL specialization out of the University of Alberta, hosted on coursera: <a href="https:&#x2F;&#x2F;www.coursera.org&#x2F;specializations&#x2F;reinforcement-learning" rel="nofollow">https:&#x2F;&#x2F;www.coursera.org&#x2F;specializations&#x2F;reinforcement-learn...</a> All courses in the specialization are free to audit.<p>For those unaware, the university of Alberta is Rich Sutton&#x27;s home institution, and he approves of and promotes the course.

If you are ever interested in the topic of RL, but wish to start learning the concepts on simpler algorithms and keep the &quot;deep&quot; part for later, I maintain a library that has most of the same design goals:<p><a href="https:&#x2F;&#x2F;github.com&#x2F;Svalorzen&#x2F;AI-Toolbox" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;Svalorzen&#x2F;AI-Toolbox</a><p>Each algorithm is extensively commented, self-contained (aside from general utilities), and the interfaces are as similar as I could make them be. One of my goals is specifically to help people try out simple algorithms so they can inspect and understand what is happening, before trying out more powerful but less transparent algorithms.<p>I&#x27;d be happy to receive feedback on accessibility, presentation, docs or even more algorithms that you&#x27;d like to see implemented (or even general questions on how things work).

Asking for the benefit of me and others since this is on the front page now - are there any resources this comprehensive for any other field of study? This guide is amazing and I&#x27;ve failed to find anything else like it. I was specifically interested in biotech (from the perspective of a software developer, i.e. practically zero biology background), but will take what I can get

If you want to play around with Spinning Up in a Docker container, then make sure you git clone the repository, then pip install -e repository. For whatever reason, if you try to directly install it with pip, it doesn&#x27;t work, at least last time I tried it. Here&#x27;s a Dockerfile and docker-compose.yaml I created some time ago: <a href="https:&#x2F;&#x2F;github.com&#x2F;joosephook&#x2F;spinningup-dockerfile" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;joosephook&#x2F;spinningup-dockerfile</a>

Can anyone recommend some less opinionated introductory resources to learn reinforcement learning that focus on first principles?

RL, including contextual bandits, is becoming more popular for personalization, i.e. adapting some system to the preferences of (groups of) individuals.<p>Plug&#x2F;Source: I did a lit. review on this topic <a href="https:&#x2F;&#x2F;doi.org&#x2F;10.3233&#x2F;DS-200028" rel="nofollow">https:&#x2F;&#x2F;doi.org&#x2F;10.3233&#x2F;DS-200028</a>

I enormously appreciate the resources OpenAI provides to start out in DRL such as this one. However, OpenAI has (purposely?) left out the brittleness of their algorithms to parameter choice and code-level optimizations [1] in the past. As a researcher myself, I would be more than surprised to hear that OpenAI did not explore this behaviour themselves. Instead, my guess would be that these &quot;inconveniences&quot; would do harm to the Marketing of OpenAI and its algos. Such deeds are far more harmful to proper understanding of DRL and applications than a nice UI is beneficial imo.<p>[1]<a href="https:&#x2F;&#x2F;gradientscience.org&#x2F;policy_gradients_pt1&#x2F;" rel="nofollow">https:&#x2F;&#x2F;gradientscience.org&#x2F;policy_gradients_pt1&#x2F;</a>

There was a discussion on r&#x2F;datascience this weekend about if anyone uses RL. Almost no one does.<p><a href="https:&#x2F;&#x2F;www.reddit.com&#x2F;r&#x2F;datascience&#x2F;comments&#x2F;iav3lv&#x2F;how_often_do_you_guys_use_reinforcement_learning&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.reddit.com&#x2F;r&#x2F;datascience&#x2F;comments&#x2F;iav3lv&#x2F;how_oft...</a>

&quot;Pray, who is the candidate&#x27;s tailor?&quot; -Hilbert<p>Who is responsible for OpenAI&#x27;s UI&#x2F;UX design? It is immaculate and should be the standard for the community. I&#x27;m always dazzled by the impeccable standards of OpenAI with regards to tone, presentation, accessibility.<p>The documentation is both familiar but distinct, an impressive achievement!<p>I have my own personal qualms on OpenAI&#x27;s ethics and virtues but am nevertheless impressed by their aesthetics and regard for their publicity. It&#x27;s always delightful to look at their work.<p>OpenAI has in my opinion, the most appropriate presentation for their ideas with marketing and branding. It feels exquisitely simple to grasp what goes on here.<p>I feel comfortable saying that the biggest obstacle in progress for AI is UI but projects such as this give me hope.