Machine Learning Lecture 27 "Gaussian Processes II / KD-Trees / Ball-Trees" -Cornell CS4780 SP17

แชร์
ฝัง
  • เผยแพร่เมื่อ 29 ธ.ค. 2024

ความคิดเห็น •

  • @bharasiva96
    @bharasiva96 4 ปีที่แล้ว +29

    KD-Trees begins at 28:50

  • @mlst3rg
    @mlst3rg 5 ปีที่แล้ว +33

    this series is a work of art. needs way more views.

  • @yuanchia-hung8613
    @yuanchia-hung8613 4 ปีที่แล้ว +15

    The best explanation for Gaussian Process ever!

  • @rajupowers
    @rajupowers 5 ปีที่แล้ว +10

    Most intuitive explanation of the topics in classroom

  • @clementpeng
    @clementpeng 4 ปีที่แล้ว +8

    Love this. Probably the clearest explanation i have seen on GP online.

  • @raedbouslama2263
    @raedbouslama2263 3 ปีที่แล้ว +5

    The previous video and the current one are the best material I watched on Gaussian Processes! Wonderful :)

    • @peterhojnos6705
      @peterhojnos6705 3 ปีที่แล้ว

      definitely! I saw many, but this one is one of the best

  • @deltasun
    @deltasun 4 ปีที่แล้ว +5

    thank you very much! I've tried a couple of times to understand GPs, but always gave up. Now i think they're much clearer to me. very very greatful

  • @abhinav9561
    @abhinav9561 3 ปีที่แล้ว

    Prof Killian killin it! Thanks prof for all the lectures. This course should be the first introduction to the Machine Learning world for everyone

  • @chamaleewickrama3276
    @chamaleewickrama3276 3 ปีที่แล้ว +1

    Omg. I love this lecture material. To the point, clear and the best!

  • @saikumartadi8494
    @saikumartadi8494 5 ปีที่แล้ว +4

    awesome simulation of a beautiful application !

  • @isaacbuitrago2370
    @isaacbuitrago2370 4 ปีที่แล้ว +2

    You make it look easy ! Thanks for the clear explanation of GP.

  • @Illinoise888
    @Illinoise888 4 ปีที่แล้ว +3

    This helps me with my exam preparation, thank you.

  • @TeoChristopher
    @TeoChristopher 4 ปีที่แล้ว +2

    To Clarify, for 26:19 , for a Gaussian Process, each data point on the X-axis would we a queried test point , the grey region would be the standard deviation and the points that we have not "queried" would be fitted according to its respective determined distribution which it itself would be a Gaussian distribution with its own mean and s.d?

  • @kevinshao9148
    @kevinshao9148 11 หลายเดือนก่อน

    9:30, so for my test data, y_test, it has 1) its own variance, 2) n correlations with respect to all observed data y1...yn, then how to determine y_test distribution? how did you get the conclusion at 11:06? Thanks!

  • @atagomes_lncc_br
    @atagomes_lncc_br 3 ปีที่แล้ว

    Best and simplest explanation of GPR.

  • @udiibgui2136
    @udiibgui2136 3 ปีที่แล้ว +3

    Thank you for the lecture, very clear! Just one question, how does the Bayesian Optimisation already have a mapped surface?

    • @kilianweinberger698
      @kilianweinberger698  3 ปีที่แล้ว +7

      initially that is just a flat surface, which is an uninformed prior.

  • @AlexPadula
    @AlexPadula 5 ปีที่แล้ว +3

    Thank you very much, these lectures are really useful.

  • @vaaal88
    @vaaal88 4 ปีที่แล้ว +3

    this is such a great lesson. Thanks!

  • @mertkurttutan2877
    @mertkurttutan2877 2 ปีที่แล้ว

    Question: Regarding hyperparameter search via GP, I recall that the earlier steps in hyperparameter search involves determining the scale of hyperparameter. How should we determine the scale? Should we use GP for both scale and minimal value at the same scale. Or, Use grid search to determine scale and then, use GP to find the value of hyperparameter.
    Thanks for both rigorous and enjoyable lectures :)

    • @akshaygrao77
      @akshaygrao77 ปีที่แล้ว

      U keep running bayes optimization which uses gaussian processes, with more iterations it converges to smaller scales itself

  • @ayushmalik7093
    @ayushmalik7093 2 ปีที่แล้ว +1

    hi Prof
    In Bayesian Optimiser I assume that algorithm for which we are trying to find out best hyper-parameters should be costly enough otherwise it will not make any sense to use GP on top of another algo.

  • @salahghazisalaheldinataban5632
    @salahghazisalaheldinataban5632 2 ปีที่แล้ว

    Seems from your explanation that the covariance matrix is a simple kernel/distance matrix that does not take into account variable importance. (1) Does that cause any issues if there are variables that have no significant prediction value?, (2) Does it mean we have to be careful about variable selection? And (3) is there a way to incorporate feature importance in the kernel?

    • @kilianweinberger698
      @kilianweinberger698  2 ปีที่แล้ว +1

      For the linear kernel that's not an issue (as your algorithm becomes identical to linear regression where you learn a weight for each dimension), however for non-linear kernels that can indeed be a problem. One common trick is to multiply each feature dimension by a non-negative weight, and also learn these weights as part of the kernel parameters.

  • @Biesterable
    @Biesterable 6 ปีที่แล้ว +1

    Hm isn't there maybe a way to do low-dimensional egg-search (if it's a manifold there should allways be some main directions) so for the start just make it elipsoid in just one dimension and for comparing distort the room so the elipsoid you're comparing with becomes a globe hm...

  • @chenwang6684
    @chenwang6684 4 ปีที่แล้ว +2

    Awesome lecture! One question is are the projects available for public? I have found homeworks but no coding projects.

    • @kilianweinberger698
      @kilianweinberger698  4 ปีที่แล้ว +3

      Sorry, I cannot post them. The projects are still used at Cornell University, and if they were public someone would certainly post solutions somewhere and spoil all the fun. :-(

  • @imblera6571
    @imblera6571 4 ปีที่แล้ว +1

    For the hyper parameter search, wouldn't the bayesian optimization approach be more likely to get stuck at a local minimum?

    • @kilianweinberger698
      @kilianweinberger698  4 ปีที่แล้ว +4

      No, Bayesian optimization is global. The exploration component makes sure that you don’t get stuck.

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 2 หลายเดือนก่อน

    Thank you Professor !!!

  • @sarvasvarora
    @sarvasvarora 3 ปีที่แล้ว

    Living for that "YAY" 😂😂

  • @chaowang3093
    @chaowang3093 4 ปีที่แล้ว +1

    This guy is brilliantly funny.

  • @rajupowers
    @rajupowers 5 ปีที่แล้ว +1

    Important @8:00

  • @vatsan16
    @vatsan16 4 ปีที่แล้ว

    One thing I would like to ask is, "what's the catch?" The algorithms seems great but where would we not want to use GPR? Is it in situations where we would like to actually know what the function is? Or are there some situations where GPR wont work well?

    • @kilianweinberger698
      @kilianweinberger698  4 ปีที่แล้ว +4

      Well, I wouldn’t recommend them for data that is very high dimensional (e.g. bag of word vectors, or images in pixel space). Also, when features are sparse splitting along features becomes tedious and too restrictive, as almost all samples always have zeros in all dimensions.

  • @ehfo
    @ehfo 6 ปีที่แล้ว +6

    are the homeworks available for public?

    • @kilianweinberger698
      @kilianweinberger698  5 ปีที่แล้ว +9

      www.dropbox.com/s/tbxnjzk5w67u0sp/Homeworks.zip?dl=0

    • @sankalpthakuravi
      @sankalpthakuravi 4 ปีที่แล้ว +2

      Kilian Weinberger you must be an angel

  • @thecelavi
    @thecelavi 5 ปีที่แล้ว

    Is it possible to use B/B+ tree instead of simple binary tree?

  • @andreariboni4242
    @andreariboni4242 5 หลายเดือนก่อน

    dead mouse got me

  • @giraffaelll
    @giraffaelll 4 ปีที่แล้ว +2

    He clears his throat a lot

  • @LauraJoana
    @LauraJoana 3 ปีที่แล้ว

    THANKS!

  • @hassanshakeel854
    @hassanshakeel854 5 ปีที่แล้ว +1

    Are all these lectures dependent on previous ones?

  • @sandeshhegde9143
    @sandeshhegde9143 5 ปีที่แล้ว +3

    KD Tree starts from th-cam.com/video/BzHJ57QCdVo/w-d-xo.html