Just finished lecture 10 and I've come back to write a review for anyone starting. *Excellent course*. Well paced, enough examples to provide a good intuition, and taught by someone who's leading the field in applying RL to games. Thank you David and Karolina for sharing these online.
I've finished both of them, and I'd say that this one has a better and much more solid content, although the one from udacity is much more light and easy to follow, so it really depends on what you want :)
what a wonderful time to be alive!! thank god we have the opportunity to study a full module from one of the best unis in the world. taught by one of the leaders of its field
0:01 Outline Admin 1:10 About Reinforcement Learning 6:13 The Reinforcement Learning problem 22:00 Inside an RL angent 57:00 Problems within Reinforcement Learning
1:10 Admin 6:13 About Reinforcement Learning 6:22 Sits in the intersection of many fields of science: solving decision making problem in these fields. 9:10 Branches of machine learning. 9:37 Characteristics of RL: no correct answer, delayed feedback, sequence matters, agent influences environment. 12:30 Example of RL 21:57 The Reinforcement Learning Problem 22:57 Reward 27:53 Sequential Decision Making. Action 29:36 Agent & Environment. Observation 33:52 History & State: stream of actions, observations & rewards. 37:13 Environment state 40:35 Agent State 42:00 Information State (Markov State). Contains all useful information from history. 51:13 Fully observable environment 52:26 Partially observable environment 57:04 Inside an RL Agent 58:42 Policy 59:51 Value Function: prediction of the expected future reward. 1:06:29 Model: transition model, reward model. 1:08:02 Maze example to explain these 3 key components. 1:10:53 Taxonomy of RL agents based on these 3 key components: policy-based, value-based, actor critic (which combines both policy & values function), model-free, model-based 1:15:52 Problems within Reinforcement Learning. 1:16:14 Learning vs. Planning. partial known environment vs. fully known environment. 1:20:38 Exploration vs. Exploitation. 1:24:25 Prediction vs. Control. 1:26:42 Course Overview
1:10 Admin 6:13 About Reinforcement Learning 21:57 The Reinforcement Learning Problem 57:04 Inside an RL Agent 1:15:52 Problems within Reinforcement Learning
Wow, this is incredible. I'm currently going through Udacity and this lecture series blows their material from GT out of the water. Excellent examples, great explanation of theory, just wow. This actually helped me understand RL. THANK YOU!!!!!
@@JousefM I agree, those explanations by GT professors were confusing and less clear, the entire DS nanodegree which had ML, DL and RL was painful to watch and understand.
I just finished the course, and the people in this comment section are not exaggerating. This is one of the best courses on Reinforcement learning. Thank you very much DeepMind, for making this free and available to everyone!
This is one of the clearest and most illuminating introductions I've watched on RL and its practical applications. Really looking forward to the following instalments.
Excellent moment around 24:10 when David makes it crystal clear that there needs to be a metric to train by (better/worse) and that it's possible - and necessary - to try to come up with a scalar metric that roughly approximates success or failure in a field. When you train something to optimize for a metric, important to be clear up-front what that metric is.
Rare opportunity to listen to Christian Bale after he is finished with dealing with criminals as Batman. On a serious note, overall great series of lectures! Thanks, prof. David Silver!
Thanks a lot for the great lectures! I enjoyed watching every one of them (even #7). This is a great complement to reading Sutton/Barto and the seminal papers in RL. I remember looking at the Atari paper in the late 2013 and having hard time to understand why everyone is going completely crazy about it. A few years later the trend was absolutely clear. Reinforcement Learning is the key to push the performance of AI systems past the threshold where the humans can serve as wise supervisors to the limit when the different kinds of intelligence help each other to improve via self-play.
Take swimming as example: learning is part that you directly jump into the water and learn swimming to survive; planning is that part that before jumping into the water, you read books/instructions on how to swim (obviously sometimes planning helps, sometimes not, sometimes counter-helps).
@1:07:00. Instead of defining P_{ss'}^a and R_s^a, it's better to define p(s',r|s,a), which gives the joint probability of the new state and reward. The latter is the approach followed by the 2nd edition of Sutton&Barto's book.
For Me : David Silver is God ❤️ What a Man ! What an Explanation. One of the Greatest Minds who changed the Dynamics of RL in the past few years.Thanks Deep mind for uploading this Valuable course for free 🤍
I have never heard a punishment described as a negative reward at any point during my 71 orbits of the Sun. You can indeed learn something new every day.
Right now I am sitting in my pajamas in the comfort of my home, eating a peanut butter and jelly sandwich and I have the ability to watch an entire course about an advanced topic online for free. What a time to be alive
@@pratikd5882 Yes, I did that. The exercises were good, but I'm not an AI guy but a simple programmer. I managed to do the exercises but I think that explainations were very concise. So in 15 minutes they explain what you get in 1 hour on those lectures. I think that is very summarized. But it's good they have exercises. So I don't think after doing that I'm actually able to do much
Do you have any suggestions about which one to start with , the Lecture series here or the RL specialization by Alberta University (on Coursera). I need to apply RL on my own project work. By the way I did the course on Machine learning by NG Andrews and I could follow the pace it was good enough for me and besides the programming exercises helped me alot than I could imagine. But I am not sure if so would be the case with RL by Coursera as well. Can you guide me on this?
The above comment was meant to be in the context of your life. Your brain is a cumulative of all your prior experiences and the choices/decisions which you make will be an a action taken by your brain(which is a markov state). So what I perceived from that statement was that, "you need to forget your past and move on".
Hi David, Thank you very much for such great video course. I would be very much appreciated if you help clear the following: 1. Is there a need for training by playing (hundreds of time) say the game as an example before being ready? if you have generalized program designed for reinforcement learning is it ready to play a new game who rules are different or need to be trained again. 2. There is mathematical proof that intermediate rewards can be calculated based on past experience - there is no need to know the intermediate rewards,. It can be derived say using deep mind Q function.- please correct - thank you. 3. In helicopter example current state depends on the past states like velocity acceleration - so not Markov 4. Can the reinforcement learning be realistically used to drive a car or run robot using generalized program that has inputs setup appropriately? Thank you, Sam
1 Hundreds of times is absolutely nothing in ML or RL. In order to learn to play games with a rich state space like go or chess, AlphaGo or AlphaZero can play billions of games. Also, by "generalised program" you mean architecture or algorithm, like algorithms based on Q-learning, concrete policy gradient methods, etc. If the rules are different, you can use the same algorithm if you reckon it is suitable, but you would have to train from scratch (unless there is a lot of similarity between the games and you can reuse some weights, as one can do with word embeddings in diverse NLP problems). 2 You are not being rigorous about the statement of your question, so it is very hard to understand what you exactly mean. It absolutely necessary to measure intermediate rewards. 3 How is the helicopter state not Markov? The helicopter movement is simulated by running a controlled ODE, and a numerical solver for an ODE is Markov by definition. Of course it is Markov. 4 Self-driving cars are an example of that.
I wonder Markov property is the same as Newtonian mechanics in physics which is deterministic like Markov states. Newtonian mechanics say that if you know every momentum and acceleration of every particles existent, you can calculate/know the future 100%.
Revision) If Markov state talks about the "instant" moment in time (delta t->0), I guess you cannot figure out the velocity or momentum so it is different from Newtonian.
Important Fact ----> By the time I am writing this there are around 700K views on Lecture 1 but only 35K on Lecture 10. So only a 0.05% chance that you'll watch all the lectures. So be patient and remember why are you starting. Thanks
As far as I think RL is about searching a vast computational space using ingenious methods. These methods let us decrease our search space and make us use less computational resources than we would require otherwise. In the end, we still do not have the perfect solution but what we do have is a better solution than all our previous worked out solutions. With an increase in computational power and ingenuity of our algorithms, we can hope of approaching to that perfect solution. But RL is still far from being applied to the real-world scenarios which are infinitely more complex than Atari games or the game of go. If someone can do that then I guess they have a solution to artificial intelligence. But I think there are many obstacles and bottlenecks that one has to overcome to do that. The first one being able to provide the ability of vision and language interpretation to an RL agent. This is where deep learning can help. But we still need better feature extracting algorithms than convolution neural nets and LSTMs. By looking at these neural nets one can see a direct analogy with the human brain. Convolutional neural net is like the visual cortex and RL is like the Neocortex (the part of the brain that receives info from all the senses processes it and takes decisions based on it, provides us with the ability to behave and act intelligently, plus it also stores memories). In this sense, the recent research in AI is going in the very right direction and I am very glad to be a part of it.
anyone can confirm if this is still relevant in 2022? I would like to study RL. It seems that there is a more recent series but with a different professor on this channel.
Just finished lecture 10 and I've come back to write a review for anyone starting. *Excellent course*. Well paced, enough examples to provide a good intuition, and taught by someone who's leading the field in applying RL to games. Thank you David and Karolina for sharing these online.
I've finished both of them, and I'd say that this one has a better and much more solid content, although the one from udacity is much more light and easy to follow, so it really depends on what you want :)
This course is more thorough and for someone who is looking to make a career in Machine Learning, you should put in the work and do this course.
Thanks for the review!
One of the best , if not the best , courses I've watched!
thanks for the review
what a wonderful time to be alive!!
thank god we have the opportunity to study a full module from one of the best unis in the world. taught by one of the leaders of its field
0:01 Outline
Admin 1:10
About Reinforcement Learning 6:13
The Reinforcement Learning problem 22:00
Inside an RL angent 57:00
Problems within Reinforcement Learning
Good job. Very thankful :)
thanks
You made the world much easier! Thanks!
Problems within Reinforcement Learning 1:15:53
Thank you for this.
The complete set of 10 lectures is brilliant. David's an excellent teacher. Highly recommended!
I love that David is one of the foremost minds in Reinforcement Learning, but he can explain it in ways that even a novice can understand.
hlo, can u please suggeest roadmap for rl.
how lucky we are to have access to this kind of knowledge only with a button ! Thank you all in DeepMind public this course
I was going to say the same. Technology has really made our life easier and better in a lot of ways. But a lot of times we take it for granted.
@@BhuwanBhatta fvy5tym 🎉4ufgc🙏😎4g🔥f9f4c6v f😎j 9c
@@BhuwanBhatta ji kghkfktghjkhhiljcujfjpjkui jikskjgjpj
1:10 Admin
6:13 About Reinforcement Learning
6:22 Sits in the intersection of many fields of science: solving decision making problem in these fields.
9:10 Branches of machine learning.
9:37 Characteristics of RL: no correct answer, delayed feedback, sequence matters, agent influences environment.
12:30 Example of RL
21:57 The Reinforcement Learning Problem
22:57 Reward
27:53 Sequential Decision Making. Action
29:36 Agent & Environment. Observation
33:52 History & State: stream of actions, observations & rewards.
37:13 Environment state
40:35 Agent State
42:00 Information State (Markov State). Contains all useful information from history.
51:13 Fully observable environment
52:26 Partially observable environment
57:04 Inside an RL Agent
58:42 Policy
59:51 Value Function: prediction of the expected future reward.
1:06:29 Model: transition model, reward model.
1:08:02 Maze example to explain these 3 key components.
1:10:53 Taxonomy of RL agents based on these 3 key components:
policy-based, value-based, actor critic (which combines both policy & values function), model-free, model-based
1:15:52 Problems within Reinforcement Learning.
1:16:14 Learning vs. Planning. partial known environment vs. fully known environment.
1:20:38 Exploration vs. Exploitation.
1:24:25 Prediction vs. Control.
1:26:42 Course Overview
I am a simple man. I see a great course, I press like
Just finished lecture 1 and can already tell this is going to be one of the absolute best courses 👌
Mr. Silver covers in 90 minutes what most books do not in 99 pages. Cheers and thanks!
1:10 Admin
6:13 About Reinforcement Learning
21:57 The Reinforcement Learning Problem
57:04 Inside an RL Agent
1:15:52 Problems within Reinforcement Learning
bless u :)
Wow, this is incredible. I'm currently going through Udacity and this lecture series blows their material from GT out of the water. Excellent examples, great explanation of theory, just wow. This actually helped me understand RL. THANK YOU!!!!!
How do you find the RL course from Udacity? Thinking about doing it after the DL Nanodegree.
@@JousefM I agree, those explanations by GT professors were confusing and less clear, the entire DS nanodegree which had ML, DL and RL was painful to watch and understand.
I just finished the course, and the people in this comment section are not exaggerating. This is one of the best courses on Reinforcement learning. Thank you very much DeepMind, for making this free and available to everyone!
Thank you so much for repeating the questions each time.
David is awesome at explaining a complex topic!. Great lecture. The examples really helped in understanding the concepts..
I am taking a Deep Learning course rn but seriously the comments here are motivating me to get into this one right away.
This is one of the clearest and most illuminating introductions I've watched on RL and its practical applications. Really looking forward to the following instalments.
I'm really appreciating the intuitive style of this course, as contrasted to the Stanford course.
Excellent moment around 24:10 when David makes it crystal clear that there needs to be a metric to train by (better/worse) and that it's possible - and necessary - to try to come up with a scalar metric that roughly approximates success or failure in a field. When you train something to optimize for a metric, important to be clear up-front what that metric is.
Rare opportunity to listen to Christian Bale after he is finished with dealing with criminals as Batman.
On a serious note, overall great series of lectures! Thanks, prof. David Silver!
the legend of all RL courses
Just completed Coursera's Reinforcement Learning Specialization and this is a nice addition to reinforce the concept I am learning.
Thanks a lot for the great lectures! I enjoyed watching every one of them (even #7). This is a great complement to reading Sutton/Barto and the seminal papers in RL.
I remember looking at the Atari paper in the late 2013 and having hard time to understand why everyone is going completely crazy about it. A few years later the trend was absolutely clear. Reinforcement Learning is the key to push the performance of AI systems past the threshold where the humans can serve as wise supervisors to the limit when the different kinds of intelligence help each other to improve via self-play.
Thank you for sharing your knowledge online. Enjoying your videos, and loving every minute of it.
Take swimming as example: learning is part that you directly jump into the water and learn swimming to survive; planning is that part that before jumping into the water, you read books/instructions on how to swim (obviously sometimes planning helps, sometimes not, sometimes counter-helps).
@1:07:00. Instead of defining P_{ss'}^a and R_s^a, it's better to define p(s',r|s,a), which gives the joint probability of the new state and reward. The latter is the approach followed by the 2nd edition of Sutton&Barto's book.
This is a superb course on so many levels. Thank you
I took this playlist as a reference for my thesis in "RL for green radio".
His name should be David Gold or Platinum I dunno. Best intro to RL on YT, thank you!
For Me : David Silver is God ❤️ What a Man ! What an Explanation. One of the Greatest Minds who changed the Dynamics of RL in the past few years.Thanks Deep mind for uploading this Valuable course for free 🤍
I actually prefer this 2015 class over the most recent 2019 one. Nothing wrong on the other expositor, but David kinda makes the course more smoothly.
I have never heard a punishment described as a negative reward at any point during my 71 orbits of the Sun. You can indeed learn something new every day.
Right now I am sitting in my pajamas in the comfort of my home, eating a peanut butter and jelly sandwich and I have the ability to watch an entire course about an advanced topic online for free. What a time to be alive
It's really nice that he gives examples.
David is awesome at explaining a complex topic!
This is really the best RL course I have seen!
The lecturer is great!
I can say that I 've found a treasure..really
Im back again, watching the whole video again.
Really love this video series. Watching it for the fifth time:)
RL could be the killer app in ML. Nice lectures to bring people up to speed!
That was brilliant. Really helping me to get my head around the subject. Thanks David
Superb David - you are one of the giants I am standing on to see a little further - thank you
The UCL lecturer is awesome. Thx for the excellent course.
lecture starts at 6:30
5 days to train per game. now is 5 minutes to complete a train based on recent papers. envolve fast!
This is my first taste of deep mind
awesome course and awesome teacher
Can you enable the automatic captioning to this content?
Great lecture, great starting point. Helped me to understand the basics of Reinforcement Learning. Thanks for great content.
Thank you for sharing.It kinda inspires me to always remember that I have to pass it on too.
just amazing and different than any intro to RL
Thank you very much, I recently got a good grade in RL thanks to your great teaching skills!!
Could you please turn on the auto generated subscript?
Plz..
As a learner of reinforcement learning, you should become an agent and improve yourself with getting rewards in this environment
Silver is a boss.
Amazing teacher I wish I could partecipate to this course! I did a course on Coursera but it was so quick to explain very complex things.
Are you referring to the RL specialization by Alberta university? If so, then how good was it on the programming/practical aspects?
@@pratikd5882 Yes, I did that. The exercises were good, but I'm not an AI guy but a simple programmer. I managed to do the exercises but I think that explainations were very concise. So in 15 minutes they explain what you get in 1 hour on those lectures. I think that is very summarized. But it's good they have exercises. So I don't think after doing that I'm actually able to do much
Do you have any suggestions about which one to start with , the Lecture series here or the RL specialization by Alberta University (on Coursera). I need to apply RL on my own project work. By the way I did the course on Machine learning by NG Andrews and I could follow the pace it was good enough for me and besides the programming exercises helped me alot than I could imagine. But I am not sure if so would be the case with RL by Coursera as well. Can you guide me on this?
42:00 - 47:55 : Information State/Markov State
57:13 RL Agent
Silver is Gold!
Great lecture. The examples really helped in understanding the concepts.
I have to admit, david silver is slightly smarter than me.
Thanks for a great lecture. I got grasp the point of reinforcement learning !
46:20 this also means that it doesn't matter how you got into this state, it will always mean the same.
Awesome! This is succinct and clarifies some concepts that I was confused of.
Any observation and reward could be wrapped up into abstract data structure in an object for sorting.
Amazing video, a very well-designed and well-delivered lecture! I'm going to enjoy this course, good job! 👍
Excellent explanation. Thank you.
The future is independent of the past given the present
- David Silver
Only if it have Markov state
The above comment was meant to be in the context of your life. Your brain is a cumulative of all your prior experiences and the choices/decisions which you make will be an a action taken by your brain(which is a markov state). So what I perceived from that statement was that, "you need to forget your past and move on".
1:09:43 value function example in ternal
1:10:33 model; to the goal
13:27 reward delayed
if you can activate the subtitle from youtube, it will be great, Thanks
Excellent Indeed!
amazing introduction and very cool
A really good introduction course!! Thank you very much!!
Please add subtitles to make it more helpful for those who are from non English native countries
What a GIFT.
Thanks for the question at 19:48!
Great video. Thanks. I really needed something like this :)
This is so good!
Hi David,
Thank you very much for such great video course. I would be very much appreciated if you help clear the following:
1. Is there a need for training by playing (hundreds of time) say the game as an example before being ready? if you have generalized program designed for reinforcement learning is it ready to play a new game who rules are different or need to be trained again.
2. There is mathematical proof that intermediate rewards can be calculated based on past experience - there is no need to know the intermediate rewards,. It can be derived say using deep mind Q function.- please correct - thank you.
3. In helicopter example current state depends on the past states like velocity acceleration - so not Markov
4. Can the reinforcement learning be realistically used to drive a car or run robot using generalized program that has inputs setup appropriately?
Thank you,
Sam
1 Hundreds of times is absolutely nothing in ML or RL. In order to learn to play games with a rich state space like go or chess, AlphaGo or AlphaZero can play billions of games. Also, by "generalised program" you mean architecture or algorithm, like algorithms based on Q-learning, concrete policy gradient methods, etc. If the rules are different, you can use the same algorithm if you reckon it is suitable, but you would have to train from scratch (unless there is a lot of similarity between the games and you can reuse some weights, as one can do with word embeddings in diverse NLP problems).
2 You are not being rigorous about the statement of your question, so it is very hard to understand what you exactly mean. It absolutely necessary to measure intermediate rewards.
3 How is the helicopter state not Markov? The helicopter movement is simulated by running a controlled ODE, and a numerical solver for an ODE is Markov by definition. Of course it is Markov.
4 Self-driving cars are an example of that.
His 2021 "Reward is Enough" paper makes us agree to the Reward Hypothesis @ 24:18 . :D
Ty for sharing, Great Lecture!!
1:08:02 transition reward model
brilliant course
Captions would really help here!
The environment state(e,t) is Markov ... though it’s unknowable.
We can't thank you enough!
Watching at 0.75X speed helps to follow better.
imp point at 1:00:30
1:00:22 gamma's value is less than 1
I wonder Markov property is the same as Newtonian mechanics in physics which is deterministic like Markov states. Newtonian mechanics say that if you know every momentum and acceleration of every particles existent, you can calculate/know the future 100%.
Revision)
If Markov state talks about the "instant" moment in time (delta t->0), I guess you cannot figure out the velocity or momentum so it is different from Newtonian.
Too good :) Best analogies.
Thank you for nice lecture
Only slide that threw me off a bit was the RL taxonomy one. There was some confusion with the redundant labeling, otherwise it was a great lecture!
Lecture 1 has half a million views, 10 has 36k.
I'm surprised it's even 36k
I found these psychologically useful. Are subtitles available?
"The future is independent of the past ,given the present."
Important Fact ----> By the time I am writing this there are around 700K views on Lecture 1 but only 35K on Lecture 10. So only a 0.05% chance that you'll watch all the lectures. So be patient and remember why are you starting. Thanks
5%, not 0.05%
Thanks very much. But I need Eng.subtitle. Could you change setting of this videos? :)
As far as I think RL is about searching a vast computational space using ingenious methods. These methods let us decrease our search space and make us use less computational resources than we would require otherwise.
In the end, we still do not have the perfect solution but what we do have is a better solution than all our previous worked out solutions. With an increase in computational power and ingenuity of our algorithms, we can hope of approaching to that perfect solution.
But RL is still far from being applied to the real-world scenarios which are infinitely more complex than Atari games or the game of go. If someone can do that then I guess they have a solution to artificial intelligence. But I think there are many obstacles and bottlenecks that one has to overcome to do that.
The first one being able to provide the ability of vision and language interpretation to an RL agent. This is where deep learning can help. But we still need better feature extracting algorithms than convolution neural nets and LSTMs.
By looking at these neural nets one can see a direct analogy with the human brain. Convolutional neural net is like the visual cortex and RL is like the Neocortex (the part of the brain that receives info from all the senses processes it and takes decisions based on it, provides us with the ability to behave and act intelligently, plus it also stores memories).
In this sense, the recent research in AI is going in the very right direction and I am very glad to be a part of it.
¡Great lecture!
Thank you very much!!
Thanks so much!
anyone can confirm if this is still relevant in 2022? I would like to study RL. It seems that there is a more recent series but with a different professor on this channel.