Q-Learning: Model Free Reinforcement Learning and Temporal Difference Learning

Поділитися
Вставка
  • Опубліковано 21 лис 2024

КОМЕНТАРІ • 89

  • @thiagocesarlousadamarsola3990
    @thiagocesarlousadamarsola3990 2 роки тому +85

    I personally love the big picture perspective that Prof. Brunton always shows. Please, continue to make these high quality videos!

  • @davidelicalsi5915
    @davidelicalsi5915 2 роки тому +13

    Professor I must sincerely thank you for the astonishing quality of this video. You were able to clearly explain an advanced concept without simplifying, going into the details and providing brilliant insights. Also I sincerely thank you for saving my GPA from my R.L. exam 😆

  • @anlehoang7030
    @anlehoang7030 9 місяців тому +3

    More casual example for TD-learning:
    Imagine a curious robot exploring a maze, searching for a hidden treasure. Unlike other methods that wait until it finds the treasure to learn, TD learning is all about learning on the fly. It uses what it already knows (like the estimated value of different paths) and immediate feedback (rewards) to improve its predictions about future moves.
    - The robot keeps track of a Q-value (Q(s_t, a_t)) for each path, which tells it how good it thinks that path is based on its past experiences.
    - When it takes a path and gets a Q-value (or reward) (like finding a clue), it compares that reward to what it expected (based on the Q-value - r_t + \gamma Q(s_{t + 1}, a_{t + 1})). This difference is called the prediction error.
    - If the reward is better than expected (positive error, or r_t + \gamma V^{old}(s_{t + 1}) - V^{old}(s_t) > 0), the robot increases the Q-value for that path, making it seem more attractive next time.
    - If the reward is worse than expected (negative error, or r_t + \gamma V^{old}(s_{t + 1}) - V^{old}(s_t) < 0), the robot decreases the Q-value, steering it away from less promising paths.

  • @jashwantraj2987
    @jashwantraj2987 Рік тому +2

    Prof. Burton, you are amazing. I never expected someone to take so much of time to explain a concept about TD. I'm one of the few people who hate reading text books to understand concepts. I rather see a video or learn about it from class.
    Thanks a lot

  • @kalimantros845
    @kalimantros845 2 роки тому +1

    I was hoping that your next video would have been about Q-learning, and here it comes!

  • @complexobjects
    @complexobjects 2 роки тому +1

    I do like the description of Q Learning. I had come up with another analogy for why it makes sense. If you took the action of going out to a party, and then happened to make some mistakes while there, we wouldn't want to say "you should never go out again." We'd want to reinforce the action of going out based on the best* possible outcome of that night, not the suboptimal action that was taken once there.

  • @usonian11
    @usonian11 2 роки тому +12

    Thank you for the outstanding production quality and content of these lectures! I especially enjoy the structure diagram organizing the different RL methods.

  • @OmerBoehm
    @OmerBoehm 2 роки тому +2

    Thank you dear Prof Brunton for this outstanding lecture. The detailed explanations and focus on subtleties are so important , Looking forward to your next videos.

  • @imanmossavat9383
    @imanmossavat9383 Рік тому +1

    I enjoy your talks. They are very clear and well structured and have the right level of detail. Thank you,

  • @antimon40
    @antimon40 2 роки тому

    Somehow I find that the explanations given by Prof. Brunton are easier to understand than those provided by video lectures from Stanford (which are also available on UA-cam).

  • @TheFitsome
    @TheFitsome 2 роки тому

    This is the best RL tutorial on the internet.

  • @areebayubi5469
    @areebayubi5469 2 роки тому +2

    Thank you so much for using very relevant analogies and very clear explanations. I think I have a much better grasp of the concepts behind Temporal Difference learning now.

  • @haotianhang3997
    @haotianhang3997 2 роки тому +3

    Thank you! It's a great video. My understanding in TD learning was deepened a lot.

  • @TheSinashah
    @TheSinashah 10 місяців тому +1

    CS PhD student here. This video provides such amazing content. Highly recommended.

  • @marzs.szzzzz
    @marzs.szzzzz 2 роки тому +3

    These are fantastic lectures, I use these as an alternative explaination to David Silvers DeepmindxUCL 2015 lectures on the same topic, the different perspective really suits how my brain understands RL. Thank you!!

  • @denchen1950
    @denchen1950 2 роки тому +1

    The video quality is incredible lol and all the concept is discussed extremely clear OMG!!
    Brilliant masterpiece bro KEEP GOING !!

  • @caiolp4
    @caiolp4 2 роки тому +6

    Great lecture! Concerning the difference between SARSA and Q-Learning, I didn't get the emphasis on Q-Learning being better for exploration. In principle, one can choose a epsilon-greedy for both methods. As a matter of fact, the SARSA method is defined in Sutton's book with an epsilon-greedy policy. I get the point that the TD target of Q-Learning does not depend on the policy itself and, therefore, is called an off-policy method. However, if one can choose a exploratory policy (e.g., epsilon-greedy) for both methods, why would SARSA be safer or less exploratory?

    • @duncanw9901
      @duncanw9901 2 роки тому

      I got the impression he was asserting that the updates to the quality function can/will/often become an undesireable feedback loop when non-optimized states are used, and I would infer that means the training steps done on those states would have an undesireably high probability of entering such states.
      What you said does seem to be convincing evidence otherwise though.

    • @jorge-george6958
      @jorge-george6958 2 роки тому

      That is a correct comment. The difference between the two lies in the Q-function updates. The way you choose your action is orthogonal (and can be more/less exploratory in either method). Also, from the video, Q-learning comes off as "better" method than SARSA, at least in problems where you don't need safe exploration, which is not accurate. It's more like a trade-off, where no method is clearly better.
      I love your videos in general. I think though, that this particular one needs a bit of a revision. Hope you don't see this as a critique, but rather as constructive feedback.

  • @BoltzmannVoid
    @BoltzmannVoid 2 роки тому +3

    this was the best explanation ever!
    thank you so much, professor!

  • @farhadebrahimzadeh3420
    @farhadebrahimzadeh3420 Рік тому

    Thank you for the clear picture. It was really well explained and others already mentioned, now I can say that I understand these techniques quite fairly well. 🙏

  • @cruise0101
    @cruise0101 2 роки тому +1

    Excellent class! Extremely easy to understand!

  • @prateekcaire4193
    @prateekcaire4193 11 місяців тому

    Thanks a lot. Not just math but also the intuition that i was looking for

  • @codypappa1667
    @codypappa1667 6 місяців тому

    Great videos. Definitely keep your face in the thumbnails. Even if the channel name doesn't ring a bell, sometimes I find your videos by just the thumbnails and always enjoy them

  • @stuartferguson11
    @stuartferguson11 Рік тому

    This whole series was good, but this one pushed me past my confusion. My neural network finally learned to play tic-tac-toe!

  • @FRANKONATOR123
    @FRANKONATOR123 2 роки тому +2

    Hi Prof. Brunton. Great vídeo as always! Please keep producing quality ML content

  • @krullebolalex
    @krullebolalex 2 роки тому +3

    Thanks a bundle Steve, this was really well explained!

  • @ajaykumar-rh2gz
    @ajaykumar-rh2gz 2 роки тому +4

    Hi Steve, Thanks for Amazing! lecture. I think the mail challenge in RL is designing our own custom environment (Multiple states and actions). It will be a great help if you can upload some lecture, suggest some link to do this job. Other comments are also welcome. Currently, I am doing some experiment on Retail pricing optimization using offline data. Looking forward.

  • @multiversityx
    @multiversityx 2 роки тому

    You gave the best explanations I've ever seen!

  • @台中人-j1v
    @台中人-j1v 2 роки тому

    I very like your descriptions about deep learning

  • @nnanyereugoemmanuel8327
    @nnanyereugoemmanuel8327 5 місяців тому

    I'm so much very grateful for these videos you make. Keep on the good work.

  • @nahuelpiguillem2949
    @nahuelpiguillem2949 Рік тому

    Thanksssss steveeee. I couldnt understand nothing before this video.Thanks again

  • @martinschulze5399
    @martinschulze5399 2 роки тому

    being a phd student of you must be a gift :D

  • @mariogalindoq
    @mariogalindoq 2 роки тому +3

    Steve: thank you again. I appreciate your work. Trying to help, let me say that I believe there is a small typo, at minute 5:29 you wrote π(s,a) = argmax_a Q(s,a), should it be written π(s) = argmax_a Q(s,a)? Also, at time 22:35, the first equation has a sum over k, should it be over n? Anyway, this is a very good video.

  • @polinagrinko1678
    @polinagrinko1678 9 місяців тому

    brilliant explanation

  • @JustinMasayda
    @JustinMasayda 2 роки тому

    One thing that seems to be either an error or just inconsistent notation is the use of TD(N) to mean an N-step TD. It seems like the value in the parentheses is supposed the value of lambda, not the number of steps of TD. TD(0) apparently should be read as, "N-step TD when lambda = 0, " while TD(1) means, "N-step TD when lambda = 1." I'm basing this off of the book "Reinforcement Learning: An Introduction - Second edition" by Sutton and Barto.

  • @alirezatavakoli7325
    @alirezatavakoli7325 6 місяців тому

    The videos are wonderful! Thank you, professor.

  • @David-nw6rz
    @David-nw6rz 2 роки тому +1

    Great lecture, but I guess your definition of on/off policy is different from the definition of Sutton/Barto. On policy doesn't necessarily mean you always take the optimal action. "[on-policy] learns action values not for the optimal policy, but for a near-optimal policy that still explores" [excerpt from a different chapter but also valid for TD].
    SARSA usually still follows a epsilon-greedy strategy.

  • @jicabe577
    @jicabe577 Рік тому

    Thanks a lot, Prof. Brunton!

  • @Chetan_Hansraj
    @Chetan_Hansraj Рік тому

    woww thank you , so well explained with lot of patience .. god bless

  • @maria4880
    @maria4880 11 місяців тому

    Thank you so much for these lectures sir!

  • @vamsimanoharreddy1468
    @vamsimanoharreddy1468 Рік тому

    Excellent Explaination

  • @r.d.7575
    @r.d.7575 2 роки тому

    Read the chapter, and I've been waiting for this video for a while. Happy to know I'm the first to comment :) Thanks, Steve. Can you explain more the bias (and especially the variance) in RL in a later video ?

  • @Julsten3107
    @Julsten3107 Рік тому

    Such a great video, thank you!

  • @anirudhthatipelli8765
    @anirudhthatipelli8765 Рік тому

    Thanks for the fantastic explanation!

  • @Throwingness
    @Throwingness 2 роки тому

    Loud and clear.

  • @manmeetworld
    @manmeetworld 22 дні тому

    @12:55 if pi were not made optimal the game would converge slowly the agent doesn't explore enough and acting outcomes become bounded in a local optimum with low next step rewards.

  • @jimlbeaver
    @jimlbeaver 2 роки тому +1

    Great explanation..very clear!

  • @yuktikaura
    @yuktikaura 2 роки тому

    Very well explained.

  • @51nibbler
    @51nibbler 2 роки тому

    ty 4 great explain greeze from switzerland :)

  • @faqeerhasnain
    @faqeerhasnain 10 місяців тому

    In value able Content.. Cant thank enough.

  • @bevansmith3210
    @bevansmith3210 2 роки тому +1

    For MC you only get the reward at the end and then divide it up among all the states. But for TD, if you are only taking one step forward, where does the reward come from? A little confused here.

  • @stevenchiu8560
    @stevenchiu8560 Рік тому

    Could you please explain it with some examples, that will be really helpful to understand these formulas, thanks!

  • @linyidai9076
    @linyidai9076 Рік тому

    Help a lot with my AI course final!!!!

  • @Pedritox0953
    @Pedritox0953 2 роки тому

    Great video! Would be nice a simple example

  • @miminh98
    @miminh98 2 роки тому +1

    Hello :) thank you for the video. I have a small question. I don't fully understand the difference between TD(0) and SARSA. Indeed, if SARSA uses the optimal action 'a' at each time step 'k', doesn't the Q-function in SARSA equal the Value function in TD(0) ? Or was there an error in my understanding ? Can you please help me see more clearly ? :)

  • @BipinOli90
    @BipinOli90 Рік тому

    Brilliant! Thanks 🙏

  • @XandreClementsmith
    @XandreClementsmith Рік тому

    In the Monte Carlo method, you have the reward discounted by gamma. Why do you discount a reward function for the entire episode?
    Furthermore, 1/n(R) would not the average reward for any step k.

  • @dbracale
    @dbracale Рік тому

    Nice talk al always. A question: at minute 5:00, does pi depend on a? It shoud not, right? The same holds for the previous video.

  • @sunaxes
    @sunaxes Рік тому

    For TD0, Why not say we do alpha x the new value (s_k+1) + (1 - alpha) x the old value (s_k). It's a very basic update method...

  • @preston748159263
    @preston748159263 Рік тому

    I would like to know what is being used to project the equations. It is apparently not video editing because he points directly to them.

  • @farhadebrahimzadeh3420
    @farhadebrahimzadeh3420 Рік тому

    there is a question that pops up in my head; if SARSA is an on-policy method, then is it OK to use e-greedy algorithm in SARSA? as you mentioned it always take into account taking the safe and on-policy action rather than random and off-policy actions?

  • @somethingirreversib
    @somethingirreversib 2 роки тому

    Great lecture!

  • @anupamadhikari139
    @anupamadhikari139 2 роки тому

    Can you explain how off policy q learning can take a suboptimal route sometimes if you are always taking the max of the actions presented?

  • @chuanjiang6931
    @chuanjiang6931 Рік тому

    One question, in terms of updating the Q function using the observed(real) reward at state k + 1, how do we know the observed(real) reward at state k + 1 since it is one timestamp in future?

  • @JustinMasayda
    @JustinMasayda 2 роки тому

    22:35 As someone else mentioned, the first equation has a sum over k, shouldn't it be over n?

  • @dam-ib9fs
    @dam-ib9fs Рік тому

    very useful

  • @titaniumsheepdog
    @titaniumsheepdog 2 роки тому

    what is the reward for r_k if we don't receive a reward after every action? Is it just assumed zero and the value is based only off the quality function?

  • @tearistovic
    @tearistovic 10 місяців тому

    Thank youuu !

  • @tubege
    @tubege 2 роки тому

    Pi of s,a should only be a function of s since the RHS calculates a. What am I missing?

  • @chymoney1
    @chymoney1 2 роки тому

    Great stuff

  • @rev0cdevs38
    @rev0cdevs38 2 роки тому

    I cannot access the new chapter in the 2nd edition. Has anybody accessed the link?

  • @VladimirGusev-m4e
    @VladimirGusev-m4e Рік тому

    You are the best:))))))

  • @dominic_lee
    @dominic_lee Рік тому

    wow, nice

  • @samirelzein1095
    @samirelzein1095 2 роки тому

    top prof!

  • @alexu176
    @alexu176 4 місяці тому

    Did not understand the TD part

  • @__--JY-Moe--__
    @__--JY-Moe--__ 2 роки тому

    👍

  • @Throwingness
    @Throwingness 2 роки тому

    A+

  • @dihancheng952
    @dihancheng952 10 місяців тому

    I don't think the comparison of q learning and sarsa is accurate.

  • @mayfields5092
    @mayfields5092 2 роки тому

    isnt q-learning, learning to play master yi

  • @djsocialanxiety1664
    @djsocialanxiety1664 8 місяців тому

    too much talk too few examples

  • @npr1m991
    @npr1m991 2 роки тому +1

    This is amazing content.
    I just have a question (still struggle with the concept of on and off policy)... at 30:18, the max Q (in Q-learning) .. which Q is it old or new ?

  • @robert-dr8569
    @robert-dr8569 Рік тому +7

    Instead of using so many words to explain, why couldn't you just use a couple of examples to explain the relationship between Q(s, a) new vs Q(s', a')? It would be so easy to understand through examples.

    • @sharannagarajan4089
      @sharannagarajan4089 10 місяців тому +1

      It’s a theoretical formula.

    • @stevewu1920
      @stevewu1920 10 місяців тому +10

      Instead of using so many words to complain, why couldn’t you just make a video illustrating these yourself?

    • @romxpl4885
      @romxpl4885 8 місяців тому

      @@stevewu1920worst take ever

    • @qadoosi09
      @qadoosi09 8 днів тому

      Agreed