Kilian Weinberger
Kilian Weinberger
  • Видео 41
  • Просмотров 1 523 807
The DBSCAN Clustering Algorithm Explained
DBSCAN has become one of my favorite clustering algorithms.
The original paper is here: www.dbs.ifi.lmu.de/Publikationen/Papers/KDD-96.final.frame.pdf
(This video is part of the CS4780 Machine Learning Class.)
Просмотров: 4 015

Видео

CS4780 Transformers (additional lecture 2023)
Просмотров 7 тыс.Год назад
A brief explanation of the Transformer Architecture used in GPT-3, ChatGPT for language modelling. (Uploaded here, for those who missed class due to the unusually nice weather :-) )
On the Importance of Deconstruction in Machine Learning Research
Просмотров 7 тыс.4 года назад
This is a talk I gave in December 2020 at the NeurIPS Retrospective Workshop. I explain why it is so important to carefully analyze your own research contributions through the story of 3 recent publications from my research group at Cornell University. In all three cases did we first invent something far more complicated, only to realize that the gains could be attributed to something far simpl...
Machine Learning Lecture 18 "Review Lecture II" -Cornell CS4780 SP17
Просмотров 12 тыс.5 лет назад
Machine Learning Lecture 18 "Review Lecture II" -Cornell CS4780 SP17
In-class Kaggle Competition in less than 5 Minutes
Просмотров 12 тыс.6 лет назад
The Fall 2018 version of CS4780 featured an in-class Kaggle competition. The stuents had 3 weeks to beat my submission, for which I only had 5 minutes. Some students challenged me to show a screencast of me actually training and uploading the model in time, so here you go. Happy Xmas.
Machine Learning Lecture 22 "More on Kernels" -Cornell CS4780 SP17
Просмотров 24 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote13.html www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote14.html
Machine Learning Lecture 37 "Neural Networks / Deep Learning" -Cornell CS4780 SP17
Просмотров 15 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote20.pdf
Machine Learning Lecture 36 "Neural Networks / Deep Learning Continued" -Cornell CS4780 SP17
Просмотров 14 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote20.pdf
Machine Learning Lecture 35 "Neural Networks / Deep Learning" -Cornell CS4780 SP17
Просмотров 21 тыс.6 лет назад
Machine Learning Lecture 35 "Neural Networks / Deep Learning" -Cornell CS4780 SP17
Machine Learning Lecture 34 "Boosting / Adaboost" -Cornell CS4780 SP17
Просмотров 17 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote19.html
Machine Learning Lecture 33 "Boosting Continued" -Cornell CS4780 SP17
Просмотров 17 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote19.html
Machine Learning Lecture 31 "Random Forests / Bagging" -Cornell CS4780 SP17
Просмотров 47 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote18.html If you want to take the course for credit and obtain an official certificate, there is now a revamped version (with much higher quality videos) offered through eCornell ( tinyurl.com/eCornellML ). Note, however, that eCornell does charge tuition for this version.
Machine Learning Lecture 21 "Model Selection / Kernels" -Cornell CS4780 SP17
Просмотров 28 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote11.html www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote12.html
Machine Learning Lecture 32 "Boosting" -Cornell CS4780 SP17
Просмотров 37 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote19.html
Machine Learning Lecture 30 "Bagging" -Cornell CS4780 SP17
Просмотров 26 тыс.6 лет назад
Lecture Notes: www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote18.html
Machine Learning Lecture 29 "Decision Trees / Regression Trees" -Cornell CS4780 SP17
Просмотров 46 тыс.6 лет назад
Machine Learning Lecture 29 "Decision Trees / Regression Trees" -Cornell CS4780 SP17
Machine Learning Lecture 28 "Ball Trees / Decision Trees" -Cornell CS4780 SP17
Просмотров 32 тыс.6 лет назад
Machine Learning Lecture 28 "Ball Trees / Decision Trees" -Cornell CS4780 SP17
Machine Learning Lecture 27 "Gaussian Processes II / KD-Trees / Ball-Trees" -Cornell CS4780 SP17
Просмотров 30 тыс.6 лет назад
Machine Learning Lecture 27 "Gaussian Processes II / KD-Trees / Ball-Trees" -Cornell CS4780 SP17
Machine Learning Lecture 26 "Gaussian Processes" -Cornell CS4780 SP17
Просмотров 74 тыс.6 лет назад
Machine Learning Lecture 26 "Gaussian Processes" -Cornell CS4780 SP17
Machine Learning Lecture 25 "Kernelized algorithms" -Cornell CS4780 SP17
Просмотров 15 тыс.6 лет назад
Machine Learning Lecture 25 "Kernelized algorithms" -Cornell CS4780 SP17
Machine Learning Lecture 24 "Kernel Support Vector Machine" -Cornell CS4780 SP17
Просмотров 20 тыс.6 лет назад
Machine Learning Lecture 24 "Kernel Support Vector Machine" -Cornell CS4780 SP17
Machine Learning Lecture 23 "Kernels Continued Continued" -Cornell CS4780 SP17
Просмотров 15 тыс.6 лет назад
Machine Learning Lecture 23 "Kernels Continued Continued" -Cornell CS4780 SP17
Machine Learning Lecture 20 "Model Selection / Regularization / Overfitting" -Cornell CS4780 SP17
Просмотров 22 тыс.6 лет назад
Machine Learning Lecture 20 "Model Selection / Regularization / Overfitting" -Cornell CS4780 SP17
Machine Learning Lecture 19 "Bias Variance Decomposition" -Cornell CS4780 SP17
Просмотров 50 тыс.6 лет назад
Machine Learning Lecture 19 "Bias Variance Decomposition" -Cornell CS4780 SP17
Machine Learning Lecture 17 "Regularization / Review" -Cornell CS4780 SP17
Просмотров 17 тыс.6 лет назад
Machine Learning Lecture 17 "Regularization / Review" -Cornell CS4780 SP17
Machine Learning Lecture 16 "Empirical Risk Minimization" -Cornell CS4780 SP17
Просмотров 27 тыс.6 лет назад
Machine Learning Lecture 16 "Empirical Risk Minimization" -Cornell CS4780 SP17
Machine Learning Lecture 15 "(Linear) Support Vector Machines continued" -Cornell CS4780 SP17
Просмотров 27 тыс.6 лет назад
Machine Learning Lecture 15 "(Linear) Support Vector Machines continued" -Cornell CS4780 SP17
Machine Learning Lecture 14 "(Linear) Support Vector Machines" -Cornell CS4780 SP17
Просмотров 45 тыс.6 лет назад
Machine Learning Lecture 14 "(Linear) Support Vector Machines" -Cornell CS4780 SP17
Machine Learning Lecture 13 "Linear / Ridge Regression" -Cornell CS4780 SP17
Просмотров 35 тыс.6 лет назад
Machine Learning Lecture 13 "Linear / Ridge Regression" -Cornell CS4780 SP17
Machine Learning Lecture 12 "Gradient Descent / Newton's Method" -Cornell CS4780 SP17
Просмотров 48 тыс.6 лет назад
Machine Learning Lecture 12 "Gradient Descent / Newton's Method" -Cornell CS4780 SP17

Комментарии

  • @sharvesh0369
    @sharvesh0369 4 дня назад

    9:50 the question asked was wrong because why would you classify the same feature again ? it produces same results so no use

  • @AyushGupta-xi1ou
    @AyushGupta-xi1ou 6 дней назад

    amazing class and students

  • @ashishsethi9878
    @ashishsethi9878 6 дней назад

    Thank you so much prof for these lectures. exam's dropbox is empty can you please reupload them?

  • @evau3376
    @evau3376 9 дней назад

    Besides super clear explanations, Professor Weinberger is hilarious. But this lecture, he took a next level as a stand up comedian! Thank you! It is just therapeutic to take your class while getting my brain rejuvenated. I can tell Professor Weinberger took lots of heart and efforts at how to make students understand by using so many good practical examples. Thank you so much being a great teacher!

  • @elching.8924
    @elching.8924 15 дней назад

    One more reminder about the importance of studying math. He is very enthusiastic, probably spent years at academia but poor math foundations reveals itself in each of his lectures.

  • @jmeliu
    @jmeliu 15 дней назад

    In the last video, the professor called an outlying point a sucker. In this class, he cited the problem of house price prediction and Bill Gates. I really like my own association.

  • @Selim-of8gq
    @Selim-of8gq 15 дней назад

    Thank you so much , you have no idea sure how much this is so helpfull <3 , Sir.

  • @jmeliu
    @jmeliu 16 дней назад

    I really suggest that all universities in the United States use Kilian's MOOC directly, but only pause during the hand-raising session for teachers to answer students' questions. If this is very damaging to self-esteem, in order to teach students well, please ask teachers from other schools to at least watch Kilian's video first.

  • @jmeliu
    @jmeliu 16 дней назад

    What if σ^2 is not a constant, for example, σ is smaller at points close to zero and larger at points farther away?

  • @preetamverma5661
    @preetamverma5661 17 дней назад

    Wow Amazing ... ❤

  • @preetamverma5661
    @preetamverma5661 17 дней назад

    Amzing ..❤

  • @preetamverma5661
    @preetamverma5661 18 дней назад

    Big thanks to the professor , world needs lot of such professors .

  • @shraddha3506
    @shraddha3506 29 дней назад

    Coolest guy ever !!

  • @TheCuriousCurator-Hindi
    @TheCuriousCurator-Hindi Месяц назад

    I have been a [senior] applied scientist at places like Amazon & Microsoft. Have been studying ML for 12 years and 7+ years industry experience at big tech. What are the chances I get into cornell PhD program? In terms of fluency I can read 80-90% PML : an introduction by Kevin Murphy in 3-4 weeks.

  • @architsharma292
    @architsharma292 Месяц назад

    goated

  • @TheCuriousCurator-Hindi
    @TheCuriousCurator-Hindi Месяц назад

    watching backwards and happy to see printer working :)

    • @SimarpreetKaur-er1dw
      @SimarpreetKaur-er1dw 9 дней назад

      hey if u are able to access the lecture notes of this course, could you pls tell?

  • @TheCuriousCurator-Hindi
    @TheCuriousCurator-Hindi Месяц назад

    I have been studying ML for last 12 years and I endorse as a learner that this course is one of the best classical ML introductory course.

  • @Learningat26
    @Learningat26 Месяц назад

    I have a prediction to make, " If the classroom has one the boards which u know move up and down, and the prof uses it over Projector, the class is going to be awesome." Thanks Dr. Kilian for the whole series.

  • @mahammadnabizade9408
    @mahammadnabizade9408 Месяц назад

    Thanks for showing the pure beauty of math underlying beneath these ML concepts. My goal is to finish this playlist by the end of 2024! Thanks a Kilian, from Azerbaijan!

  • @SrikrishnanDivakaran
    @SrikrishnanDivakaran Месяц назад

    Thanks Prof. Kilian. Extremely precise, compact and yet rigorous. Phenomenal resource not only for students but also for other researchers and practioners of ML. Enormously grateful for your efforts in sharing deep insights in an elementary and accessible manner.

  • @patrickmesana5942
    @patrickmesana5942 2 месяца назад

    Kilian is amazing

  • @Selim-of8gq
    @Selim-of8gq 2 месяца назад

    you are like a gift from Nature <3 .

  • @ruiqizhu8946
    @ruiqizhu8946 2 месяца назад

    Goat prof

  • @saajanrajak3054
    @saajanrajak3054 2 месяца назад

    5:57 - boundary point can be assigned to any one group. but it may change if we restart the dbscan model. does it affect the model?

    • @kilianweinberger698
      @kilianweinberger698 2 месяца назад

      Yes, it depends on the order in which we start. That's OK though. After all there isn't a true answer and other algorithms (e.g. K-means) are even more initialization dependent.

  • @Selim-of8gq
    @Selim-of8gq 2 месяца назад

    <3

  • @ChenLiu-nc5tg
    @ChenLiu-nc5tg 2 месяца назад

    This is cool!

  • @Lfmpereira1
    @Lfmpereira1 3 месяца назад

    Amazing lecture.

  • @MarcoGelsomini-r8c
    @MarcoGelsomini-r8c 3 месяца назад

    34:52 very interesting point!

  • @amorphous8826
    @amorphous8826 3 месяца назад

    👍

  • @szaybhattacharya1022
    @szaybhattacharya1022 3 месяца назад

    I will start my machine learning journey, I have the knowledge of stat , calculus, linear algebra.Should I follow this series?. From comments it seems this is a very very good lecture series. Or this would be too advance for a newbie like me?

  • @adosar7261
    @adosar7261 3 месяца назад

    At 39:53 the example seems a little bit off to me. I would expect Newton's method to converge in a single step (the function looks like a quadratic).

    • @leviiida9967
      @leviiida9967 2 месяца назад

      The second order Taylor's approximation gives the polynomial equation of the parabola that approximates the function at a point .Here the function is the small parabola and the approximation is the other wider parabola, Newton's method minimises the wider parabola, that's why it will just shoot off in a single step

    • @adosar7261
      @adosar7261 2 месяца назад

      @leviiida9967 Doesn't the approximation looks a bit off? I understand the purpose of the drawing, i.e. that we bounce back and forth. It just felt "too contrived", that is if we construct the second order approximation it would be steeper.

  • @amorphous8826
    @amorphous8826 3 месяца назад

    👍

  • @amorphous8826
    @amorphous8826 3 месяца назад

    👍

  • @srisaisubramanyamdavanam9912
    @srisaisubramanyamdavanam9912 3 месяца назад

    compression comparison for cross entropy is just damn good....

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks a lot. Wished I could have attended 'ML for Data Science' as well

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you Professor !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    "If you can't read my handwriting, you are not alone" :)

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you Sir !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks a lot !!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you !!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks a lot !!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks a lot Sir !!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thanks !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you Professor !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    The time watching these lectures passes in no time !! Thanks, Professor!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you for the amazing Lecture !!!

  • @vocabularybytesbypriyankgo1558
    @vocabularybytesbypriyankgo1558 3 месяца назад

    Thank you so much Sir !!!

  • @shashanksharma1498
    @shashanksharma1498 4 месяца назад

    He is a fantastic teacher. And all the energetic teaching has given him a bad throat.

  • @jandrzej1264
    @jandrzej1264 4 месяца назад

    I like your way of teaching more than Andrew or anyone else, cause they assume students or anyone watching is at least tiny intelligent, which just isnt the case. After all we are all just stupid, except for few people, or that one kid in class