- Видео 35
- Просмотров 45 666
ML New Papers
Добавлен 24 сен 2023
ML Research Paper Explained - Bet on a Coin and Achieve a Better-than-KL PAC-Bayes Bound
Link to the paper:
proceedings.mlr.press/v247/kuzborskij24a/kuzborskij24a.pdf
Interested in the fantastic topic of PAC-Bayes bounds? Why not check out the best book on the question, written by the best researcher on the topic:
User-friendly Introduction to PAC-Bayes Bounds, written by Pierre Alquier, at now publishers (Foundations and Trends in Machine Learning).
Wanna know some of the challenges of modern Large Language Models? Interested in the basics of online learning and how they can help? Link to incredible Francesco Orabona’s interview:
ruclips.net/video/_rjhTAAIefA/видео.htmlsi=IHZqDUDRKMUPwdue
Interested in concentration inequalities? Check out the video on Michel Talagrand’s work, wh...
proceedings.mlr.press/v247/kuzborskij24a/kuzborskij24a.pdf
Interested in the fantastic topic of PAC-Bayes bounds? Why not check out the best book on the question, written by the best researcher on the topic:
User-friendly Introduction to PAC-Bayes Bounds, written by Pierre Alquier, at now publishers (Foundations and Trends in Machine Learning).
Wanna know some of the challenges of modern Large Language Models? Interested in the basics of online learning and how they can help? Link to incredible Francesco Orabona’s interview:
ruclips.net/video/_rjhTAAIefA/видео.htmlsi=IHZqDUDRKMUPwdue
Interested in concentration inequalities? Check out the video on Michel Talagrand’s work, wh...
Просмотров: 378
Видео
The 2025 Guide to Machine Learning Conferences
Просмотров 45721 день назад
PDF schedule of all the 2025 machine learning conferences mentioned in this video (calendar and map): drive.google.com/file/d/1D3KamZKxAe3Cs9U42zDJefrqiBUNt-jQ/view?usp=sharing Link to the video on the Reinforcement Learning Conference (RLC): ruclips.net/video/js_zmgjGO1Y/видео.htmlsi=E7Ymo4sxzpLKKzfP In this video, we cover 17 of the most major conferences in machine learning, AI, data mining,...
Ramanujan, an Indian Self-taught Math Legend
Просмотров 835Месяц назад
December 22nd is India’s National Mathematics Day, which is Srinivasa Ramanujan’s birthdate. But who is really Ramanujan? In today’s video, you will discover his life, shaken by poverty, war and health issues, as well as some of his most famous contributions in math. We will even solve together one of his earlier math problems, published in the Journal of the Indian Mathematical Society. On thi...
Top 3 Best Books to Learn ML/AI and Become a Professional
Просмотров 4,7 тыс.Месяц назад
Check out my interview of Francesco Orabona, expert in online machine learning and pioneer of parameter-free algorithms: ruclips.net/video/_rjhTAAIefA/видео.html In this video, I share with you my top 3 best books to learn machine learning, and become a professional in the field. If you want to get into the tech industry, or quantitative finance, you cannot go wrong with those ones. They are al...
Interview of Francesco Orabona - from LEGO BUILDER to ONLINE MACHINE LEARNING EXPERT
Просмотров 7592 месяца назад
Francesco Orabona’s paper on parameter-free kernel-based SGD: proceedings.neurips.cc/paper_files/paper/2014/file/0ff8033cf9437c213ee13937b1c4c455-Paper.pdf Francesco Orabona’s paper which connects coin-betting and online learning (w. David Pal): proceedings.neurips.cc/paper_files/paper/2016/file/320722549d1751cf3f247855f937b982-Paper.pdf In this video, we have the pleasure to receive one of the...
Explanation of the Nobel Prize in Physics (and AI) 2024 - John Hopfield, Geoffrey Hinton
Просмотров 1 тыс.3 месяца назад
In this video, I explain in a hopefully nice and approachable way the work of John Hopfield and Geoffrey Hinton which owed them the Nobel Prize in Physics 2024. We also discuss about the potential dangers of AI, as well as the controversy of giving the award to their work in AI, which is not directly pure physics. On this RUclips channel, you will find: recent research papers explained in 10 mi...
Should you choose an EASY or a HARD PhD Research Project?
Просмотров 4023 месяца назад
Check out my interview of ML Researcher (Google DeepMind, UCL) Arthur Gretton: ruclips.net/video/wbVoH9QUm80/видео.htmlsi=LZDMHQnuLS8xCuZZ Check out my interview of ML Researcher (UT Austin) Amy Zhang: ruclips.net/video/NQ7chmF1MwI/видео.htmlsi=Yq-_ZsklD0Mj-VD3 Should you choose an easy or a hard PhD research topic? This is a question that many students, myself included, have asked themselves. ...
Interview of Arthur Gretton ML Researcher at Google DeepMind
Просмотров 5 тыс.4 месяца назад
In this video, we have the pleasure to receive one of the most famous researchers in machine learning of our time: Arthur Gretton. He is an expert in the fields of kernel methods, statistical tests and causality among other things. Today, he explains to us the connections between academia and the industry, as well as his take on the recent boom in AI startups based on LLMs (Large Language Model...
I am a PhD student and I do NOTHING
Просмотров 1,9 тыс.4 месяца назад
Check out my interview of Reinforcement Learning Researcher Amy Zhang: ruclips.net/video/NQ7chmF1MwI/видео.htmlsi=sLh09VMYjUeXGgzp Check out my interview of Reinforcement Learning Researcher Eugene Vinitsky during my presentation of the Reinforcement Learning Conference: ruclips.net/video/js_zmgjGO1Y/видео.htmlsi=xw9ExPYUuSI1Oqj0 Have you ever felt like you are doing nothing most of the time? M...
How to Boost your ML Skills this Summer?
Просмотров 5826 месяцев назад
In this video, I share with you how to boost your ML skills and become an AI expert by the end of the summer. #research #mlresearch #ml #machinelearning #machinelearningresearch #phd #master #ai #artificialintelligence #datascience #mlresearcher #machinelearningresearcher #statistics #optimization #math #computerscience #cs #tech #technology 0:00 Intro 0:47 Rest, Work or Study? 2:07 Practice or...
The first Edition of the new Reinforcement Learning Conference (feat. two chairs of the conference)
Просмотров 9096 месяцев назад
The first Edition of the new Reinforcement Learning Conference (feat. two chairs of the conference)
ML Research Paper Explained - A Reinforcement Learning Algorithm with no Burn in Cost
Просмотров 9997 месяцев назад
ML Research Paper Explained - A Reinforcement Learning Algorithm with no Burn in Cost
How to Find YOUR own Research Topic in ML (or any other field)
Просмотров 1,4 тыс.7 месяцев назад
How to Find YOUR own Research Topic in ML (or any other field)
CHECKLIST before applying to a Master or PhD in the US
Просмотров 6438 месяцев назад
CHECKLIST before applying to a Master or PhD in the US
Reinforcement Learning, Networking, Conferences: Interview with an ML EXPERT (feat. Amy Zhang)
Просмотров 1,3 тыс.8 месяцев назад
Reinforcement Learning, Networking, Conferences: Interview with an ML EXPERT (feat. Amy Zhang)
A Research Conference for High School Students?? (NeurIPS)
Просмотров 9599 месяцев назад
A Research Conference for High School Students?? (NeurIPS)
The Abel Prize 2024 is in machine learning? Talagrand: a bad pupil who changed ML
Просмотров 3,4 тыс.9 месяцев назад
The Abel Prize 2024 is in machine learning? Talagrand: a bad pupil who changed ML
VLOG TO THE MACHINE LEARNING SUMMER SCHOOL (OKINAWA 2024)
Просмотров 1,4 тыс.10 месяцев назад
VLOG TO THE MACHINE LEARNING SUMMER SCHOOL (OKINAWA 2024)
ML Research Paper Explained - Accelerated Optimization for Sparse PageRank
Просмотров 59010 месяцев назад
ML Research Paper Explained - Accelerated Optimization for Sparse PageRank
In-Depth Explanation of Google's Algorithm PageRank: how it sorts the web
Просмотров 71411 месяцев назад
In-Depth Explanation of Google's Algorithm PageRank: how it sorts the web
Where do Data Scientists - ML Engineers... find Resources to Learn ML?
Просмотров 1,1 тыс.11 месяцев назад
Where do Data Scientists - ML Engineers... find Resources to Learn ML?
Alpha Geometry, an AI which can solve math problems?
Просмотров 2,2 тыс.Год назад
Alpha Geometry, an AI which can solve math problems?
Best 5 Habits to become an ML EXPERT (DATA SCIENTIST, ENGINEER, RESEARCHER)
Просмотров 903Год назад
Best 5 Habits to become an ML EXPERT (DATA SCIENTIST, ENGINEER, RESEARCHER)
ML Research Paper Explained - A Best-of-both-Worlds Bandit Algorithm
Просмотров 1,3 тыс.Год назад
ML Research Paper Explained - A Best-of-both-Worlds Bandit Algorithm
ML Research Paper Explained - A Surrogate Bound via the Modulus of Convexity
Просмотров 1,3 тыс.Год назад
ML Research Paper Explained - A Surrogate Bound via the Modulus of Convexity
How to understand state of the art ML research
Просмотров 1,8 тыс.Год назад
How to understand state of the art ML research
ML Research Paper Explained - A Robust Statistical Estimation of Copula
Просмотров 2,5 тыс.Год назад
ML Research Paper Explained - A Robust Statistical Estimation of Copula
Great work! You are wery interesting content creator.
Can i know what your channel basically about what , what do you teach ?
I share my passion on machine learning and AI, and how to get into those fields. That may be via explanations of research papers like this one, but also famous researchers interviews, advice for prospective students and so on.
Very interesting and top clear explanations ! 👍
Thank you so much!
Great content... not the fluff I see on most ML videos.
Thank you very much for your comment!
Yes I am looking for my own research topic. Thanks
Thank you for your comment, I am glad to hear the video was helpful for you!
@MLNewPapers-xo5vm I wish you can help out. Thanks
You're doing an amazing job
Thank you so much for your comment!
Happy new year Charles ! All the locations seem pretty cool ! Conference traveling is nice trick to discover new places and have fun ! :)
Happy New Year Kevin! I couldn't agree more: most researchers, when they are allowed to, take advantage of conferences to enjoy new places, often with the new friends they met at conferences!
Thanks for informative video! 😊
Thank you for your comment!!
Happy New Year🎉. This video edit is very useful. Thank you.
Thank you so much, I am glad that it is useful for you and happy new year!
Very convenient !
Thank you, I really appreciate the support!
Can you make that part 2 on deep learning and optimization please? Hard to determine as a beginner what youtube videos are from genuine ml experts or just your typical data science bros who are just marketers, not real ml and deep learning practitioners. Thank you so much for what you do.
Thank you so much for your comment! I will continue this series of videos this year then!
@@MLNewPapers-xo5vm When do you think you'll be able to make Part 2? Thank you.
The books are really great! Thanks for recommendations.
Thank you very much for your comment! I am glad you like them.
Excellent reviews on the books. I am very ignorant about this subject so thank you for guiding ppl.
Thank you so much for your comment! I am glad I could help and I wish you the best in your stats and ML journey!
Charles- really enjoyable! Thanks!
Thank you very much Prof Tibshirani, I realty love your books and I am sure they will be very helpful for others as much as they have been for me!
Super pedagogical!👍 0:23
Thank you so much for your comment!!
Cheers ! From India 🇮🇳
Thank you vey much for your comment, cheers from Japan!
Wow!!!!!! Super interesting ! We feel the extraordinary research, the quality of the iconography, the tone is remarkably adapted, all this with a synthetic spirit: all this in 16 minutes with an extremely clear mathematical explanation: great work professional !👏
Thank you so much for your very supportive comment, I appreciate so much!!
Very clear and synthetic overview: very convincing. Thank you so much for sharing all of this.
Thank you so much for your comment. All the best on your ML journey!
Hey, Can you also cover how to learn generative AI from the basics and what we need to know before going down that path, like LLMS and gen AI stuff? Great content!!
Thank you for your comment, I can do that!
Great recommendations! Will add them to my holiday reading-list 😊 thanks for sharing
Thank you so much for your content, I am happy that helps!
Link to the full interview: ruclips.net/video/_rjhTAAIefA/видео.htmlsi=xu8vEdBJGD5m4Uu_
the parameter free gigachad himself
Hahaha yes, that is!
I fell in love with his research the moment I went to his talk on parameter-free kernel SGD he mentioned!
I completely agree with you: his presentations are really insightful, just like his research (soon on this channel...).
I dunno if people realize it. This is Francesco Orabona, like really one of the best researchers in the world!!!!! So glad to see him on interview, and boy, what an interview: the questions were right on point, excellent remarks!! Def the interviewer knows what he is talking about! Great job!
Thank you so much for your comment, I really appreciate it! And I def agree, Francesco is one of the best ML researchers of our era.
Very interesting dialogue and Thanks for sharing his wonderful experience
Thank you so much, I really enjoyed the experience and time with Francesco Orabona too!
Great dialogue :)!! Keep going!
Thank you so much! I really appreciate the support!
Really enjoyed your conversation! Specially the part about networking.
Thank you so much for your comment! I also learnt so much about networking when interviewing Amy Zhang, Arthur Gretton as well as the next researcher coming up soon on the channel (spoiler alert).
How do i use it
Love the approach! Your channel is inspiring doctor, thank you for your effort !
Thank you very much for your comment, I really appreciate!
Great video. One thing, it is properly pronounced “No-bell”, not “nah-bill”
Thank you very much for your comment! And I will keep that in mind if I make a related video on the Nobel Prize in the next few weeks ;)
Very interesting corelation that low productivity leads to lower work time. That really is a negative spiral that can be very demotivating. Nice to be aware of it! Thanks for pointing this out.
Thank you very much for your comment! I completely agree, this is really a vicious circle: the less one produces, the less one works, and so on.
When you work on multiple projects at a time, do all/ several of them have deadlines when you need to finish them? I find multiple deadlines often create too much stress. With multiple projects, you also have to spend mental energy planning/ deciding when you work on which ones. Switching between different complex projects can also be very exhausting as you forget how things worked or what you were working on. So doing them one after the other can be faster than doing both simultaneously. That may depend on your person though.
Thank you for your comment and very good point you have here! The projects I work on often have no specific deadline, but when I work with senior colleagues, we set together a (rough) deadline to finish the work to keep the productivity up. If I engage in several projects, I make sure the deadlines do not overlap, with (in my case) at least 1-2 months between 2 deadlines. And usually, I spend most of the time on the project which has the closest deadline. For example, if I work on 3 projects, I would spend 3 days on the project which has the earliest deadline, and 1 day on each of the others. Working at least half of the time on the project with the earliest deadline keeps the number of hours large, while working on several projects also keeps my productivity high. And as you pointed out, I always make sure to work on each project at least half a day per week in order not to forget what I did. In my case at least, what really helped me is to have a fixed schedule of what to do, so I never ask myself what to do. Besides increasing my productivity, it helps me not to spend time thinking about what to do and when to do it, which decreased my productivity and was one of the main sources of my stress. Hope that helps!
Thank you. As a PhD student struggling this is useful for me
Thank you very much for your comment, I am glad to hear that my video is helping you on your PhD journey!
Thought this had way more views than it did. Good video keep it up
Thank you so much for your comment and support, I really appreciate!!
voice is quiet, so was a little hard to hear w/out turning up. also id keep the subcribe stuff till the end, and maybe some music like what practical engineering has for the intro
Thank you so much for your advice. I will apply it to the next video!
Video very clear and very well illustrated to understand everything !
Thank you so much, I really appreciate!
Amusing 😊
Very clear and concise explanations which really help
Thank you very much for your nice comment!
Congrats for hitting +1K subscribers!!
Thank you so much!!
Vladimir Lock
Fun, friendly and interesting 😊
Thank you so much for your support!
THX super-great insightful vid all-time quote: "use what works"
Thank you so much! And yes, I agree, def a very useful advice for any ML engineer or researcher: focus on the method which works best for your problem!
Great interview! very interesting, thanks 👍
Thank you so much for your support!
So cool to have Arthur Gretton in interview and so interesting to have his points of view and advices ! Thanks a lot to ML New Papers 😊
Thank you so much for your comment! Yes, I was really happy to have him on the channel and I learnt a lot from his advice too!
Amazing and stimulating video ! So useful ! Thanks for the very valuable advice
Thank you very much for your support, I highly appreciate!
I am totally up! after watching this video, no more slump, I am 100% boosted for my project. Thanks.
Thank you for your comment. I am glad to hear that!
Last question could only be answered by those who took RL in their college.😅
Interesting
Thank you :)
Thank you
where do we get to watch RLC?