Thanks for your comment. For people new to physics (I am one of them) I can't recommend strongly enough Leonard Susskind's 'The Theoretical Minimum' courses -- http://theoreticalminimum.com/courses . I'm currently going through the one on Special Relativity and EM concurrently with Hestenes' text.
This is great, thanks so much for linking this. I acquired a taste for mathematics way too late and didn't get a chance to grasp physics at university because of it, the Theoretical Minimum courses seem to be exactly the second chance I was looking for.
If you get the first one, make sure to get the one titled 'Classical Mechanics' and not just 'The Theoretical Minimum' as the former is a more recent printing with many corrections.
I'm currently working on an RL project based on an OpenAI Gym environment and have been reviewing the different frameworks available. So far I’ve come across:
- OpenAI Baselines (more a collection of algorithms than a framework)
- Keras-RL (looked ideal but has been abandoned)
- Tensorflow Agents (An 'official'? Tensorflow library, but very basic- only one algo at present)
- rllab (Developed by OpenAI people but seems to be abandoned)
- OpenAI Lab (?)
- TensorForce
My main concerns are: 1. Soundness of the algo implementations. 2. Modularity, ease-of-use, compatibility.
I first looked at Baselines as it seemed to best address the first concern but ran into frustrations when for example the DeepQ implementation didn’t work if my Gym’s action_space was a Tuple space. I am working with a team unfamiliar with RL so want something that is as plug-n-play as possible, like Keras. So far TensorForce looks promising. Can anyone add anything more? Thanks
At least in terms of integration, TensorForce aims to be a "plug and play" library. However, RL is not at a stage right now where you can just plug an algorithm to any kind of problem and expect it to learn. Hyperparameter tuning is always necessary.
Still, TensorForce does provide pluggable implementations of state-of-the-art algorithms as well as runner utilities and environment abstractions to make it easy to connect your learning problem to it.
Try F# my man. In my experience its the best option for a statically typed functional language if you want to actually get stuff done and not just dick around. Also as you're coming from C# it would be the logical choice since the tooling is all there and you're already familiar with it. Visual Studio Code + Ionide plugin is what I use. I really don't understand why F# isn't more popular. Maybe it has something to do with its Microsoft roots making it uncool among the hip functional crowd?
I got fed up with up with Python and switched to F#. My main issue was the lack of static typing made testing and debugging a major timesink and headache.
It's early days but so far F# it ticks all my boxes and I'm finding it an extremely nice language to write in. It is just as terse as Python (due to type inference) but statically typed, and there is a great plugin Ionide for VSCode which makes for a really polished development environment.
Plan is to use Microsoft's CNTK for ML/DL stuff.
For anyone frustrated with Python's duck typing, I highly recommend you check out F#.
One I like: 'Weird games for your pleasure' http://store.steampowered.com/curator/7099409-Weird-Games-fo...