'Deep Learning'에 해당되는 글 593건

  1. 2021.02.08 안녕하세요 TensorFlow KR 여러분! Style Transfer 분야의 핵심이 되는 두 논문을 소개하는 영상을 만들어 공유합니다. 논문이나 코드 관련 질문은 이 페북 댓글로 남겨주시면 답변 드리겠습니다!Style ..
  2. 2021.02.05 [AP & mAP 내용 정리] 분류기의 성능 평가를 위한 지난 포스팅(정밀도(Precision)와 재현율(Recall) 내용 정리)에 이어 이번
  3. 2021.02.04 ML 고수분들께 질문드립니다! 딥러닝 공부를 해오면서 요즘 더욱 더 기본기의 중요성을 느끼고 있는데 기본기를 직접 구현을 통해 복습해보려고 하는
  4. 2021.02.03 Google의 머신러닝 엔지니어링 실무 지침서입니다. 머신러닝 프로젝트 구조화에 대해 공부하다가 찾게 된 문서인데, 내용이 너무 좋아 공유 드
  5. 2021.02.03 안녕하세요. Tensorflow KR 카카오브레인에서 자연어처리 라이브러리인 Pororo를 출시하게 되어서 글을 작성합니다. Pororo는 영
  6. 2021.02.01 Self-Supervised 계열의 paper를 정리하려는 목적으로 만든 repository에 꾸준히 스타가 생기고 있어서 기분이 좋아 정리를
  7. 2021.01.31 안녕하세요! NLP를 공부하시는 분들, 혹은 처음 접하는 분들께 도움이 될까 하여 올려봅니다! 🙂 빅데이터 연합동아리 투빅스에서 7주동안
  8. 2021.01.31 GPT3-small / T5-base / Bertshared-base 모델을 공개합니다. Tensorflow와 pytorch모두에서 사용이 5
  9. 2021.01.27 [self] Tensorflow 텐서플로우 2.0 괜찮은(들을) 강의 리스트
  10. 2021.01.21 안녕하세요 딥러닝 논문읽기 모임 입니다! 오늘 소개해 드릴 논문은 현재 많은 Image classification 분야에서 SOTA를 달성했던
  11. 2021.01.20 21 Resources for Learning Math for Data Science
  12. 2021.01.17 안녕하세요. 평소 머신러닝, 딥러닝에 관심이 많았는데 캐글은 이번에 처음 시작하게 되어 질문 드립니다. 1. 연구에서도 그렇지만 캐글에서는 특히
  13. 2021.01.15 "RepVGG: Making VGG-style ConvNets Great Again" 논문 소개 안녕하세요, Cognex 이호성입니다. 재미있
  14. 2021.01.14 안녕하세요. Tensorflow KR. 자연어처리 리뷰모임 [집현전]의 운영자 고현웅이라고 합니다. 이번 12월 한 달 동안 집현전에서 만든 자
  15. 2021.01.08 텐서플로우 입문 강의를 두 개 제작하여 머신러닝 야학에 공개했는데요. 열심히 만들었는데... TensorFlow KR 그룹에 공유해봅니다. :)
  16. 2021.01.04 새해부터 데이터셋 Bomb이 터졌네요. Eleuther AI 에서 공개적으로 수집 및 훈련에 사용이 가능한 영어 텍스트 825GB 코퍼스 Pil
  17. 2020.12.31 안녕하세요. 제 깃헙에 아래 첨부한 사진에 나오는 모델들을 개인 공부 차원으로 구현해 보았으니 필요하신 분들은 편하게 가져가세요~~~~ PS
  18. 2020.12.28 안녕하세요 Tensorflow KR. 이번에 제가 하고있는 논문스터디에서 기계번역 서베이를 준비하면서 자연어처리 백그라운드(Seq2Seq 부터
  19. 2020.12.28 대규모 한국어 텍스트 데이터로 학습한 pre-training 언어 모델을 공개합니다!!​ ​ 블로그, 리뷰, 댓글과 같이 사람들이 직접 사용하는
  20. 2020.12.23 안녕하세요, TensorFlow KR 여러분! 생성 모델(Generative Model)은 실제로는 존재하지 않지만, 있을법한 데이터를 만들
  21. 2020.12.22 안녕하세요. 한국어 문장 분리기로 많은 사랑을 받고 있는 Korean Sentence Splitter(KSS)의 새로운 소식을 어디다 남길까
  22. 2020.12.18 Lectures for UC Berkeley CS 285: Deep Reinforcement Learning Playlist : https://
  23. 2020.12.17 Deep Learning in Life Sciences by Massachusetts Institute of Technology (MIT)
  24. 2020.12.16 안녕하세요. Tensorflow KR. 얼마전에 SKT에서 한국어 Seq2Seq 사전학습 모델인 KoBart를 공개하였는데요. 이를 git 2
  25. 2020.12.07 안녕하세요 딥러닝 논문읽기 모임 입니다! 논문 발표자료 : https://www2.slideshare.net/taeseonryu/explai
  26. 2020.12.07 6 free, high-quality Udemy courses for you today: Python https://www.udemy.com
  27. 2020.11.25 Cheat sheet for python Neural network Machine Learning Data Science for newbies ALL PDF.https://bit.ly/394jjDH
  28. 2020.11.20 Top 13 YouTube Channels to Learn Data Science
  29. 2020.11.19 https://youtu.be/MMQjmeTmoxQ  최재식 교수님께서 설명가능한 AI(딥러닝)에 관한 여러 모델을 알기 쉽게 말씀해 주셨습니
  30. 2020.10.13 안녕하세요, TensorFlow KR 여러분! StarGAN 리뷰 및 코드 실습 영상을 만들어 공유합니다. 한 장의 사진을 업로드해서 성별/나이

 

안녕하세요 TensorFlow KR 여러분! Style Transfer 분야의 핵심이 되는 두 논문을 소개하는 영상을 만들어 공유합니다. 논문이나 코드 관련 질문은 이 페북 댓글로 남겨주시면 답변 드리겠습니다!

Style Transfer를 설명하는 영상들은 이미 많이 있지만, 하나의 영상 안에 내용부터 전체 코드까지 짜임새있게 정리된 영상은 많이 없는 것 같아서 만들어 보았습니다.

본 영상은 ① 논문 핵심 요약(PPT) ② Google Colab에서 바로 실행해 볼 수 있는 PyTorch 코드 ③ 원본 논문 리딩을 모두 포함합니다.

이 두 논문에 대한 영상은 각각 1시간가량으로 꽤 길지만, 꼼꼼히 공부해보고 싶은 분들께 도움이 될 것 같아 만들어 공유합니다.

1. Image Style Transfer Using Convolutional Neural Networks (CVPR 2016)

https://www.youtube.com/watch?v=va3e2c4uKJk

초창기 방법을 제안한 논문입니다. Style Transfer 결과 이미지의 퀄리티가 우수하지만, 속도가 오래 걸린다는 단점이 있습니다.

2. Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization (ICCV 2017)

https://www.youtube.com/watch?v=OM-6zYYRYfg

Style Transfer의 속도를 향상시키는 아키텍처를 제안한 논문입니다.더불어 임의의 스타일 이미지로부터 스타일 정보를 가져와 반영할 수 있습니다.


전체 PPT 자료 및 소스코드: https://github.com/.../Deep-Learning-Paper-Review-and...

Posted by uniqueone
,

[AP & mAP 내용 정리]
분류기의 성능 평가를 위한 지난 포스팅(정밀도(Precision)와 재현율(Recall) 내용 정리)에 이어 이번엔 객체 검출기의 성능 평가를 위한 AP와 mAP에 대한 내용을 정리해보았습니다.
https://pacientes.github.io/posts/2021/02/ml_ap_map/

Posted by uniqueone
,

ML 고수분들께 질문드립니다!
딥러닝 공부를 해오면서 요즘 더욱 더 기본기의 중요성을 느끼고 있는데 기본기를 직접 구현을 통해 복습해보려고 하는데 직접 구현해볼만한 내용들이 어떤게 있을까요??? ML 기본기를 탄탄하게 하기 위해 직접 구현해보면서 복습해보려고 합니다!! 많이 추천 부탁드릴게요!!ㅎㅎ

자신의 주위에서 흥미로운 문제를 찾아보세요. 우리나라 각 기관에서 공개한 데이터도 많습니다.
이 단계, problem finding이 제일 중요합니다. 기술이나 도구는 보편화됩니다. 갈수록.
예를 들자면 hwp, word, LaTex..등 도구를 몰라서 좋은 글을 못쓰는 것이 아니거든요.
다른 사람의 (코드,데이터) 깃헙에서 내려받아 백날 돌려봐도 큰 도움이 안됩니다.
아주 작은 데이터라도, 본인이 진짜 궁금해하는 문제부터 시작해보세요. 그래야 모형평가도
가능하고 새로운 모형으로 개선할 수 있습니다. ML에서 general solution은 없거든요.
그런 "통일장 이론"이 목표이긴 하지만, 크게보면 결국은 NP 공간의 local에 빠질 수 밖에 없죠.
요약: 내가 잘 알고있는 "나와바리"에서 일단 시작하자.

Posted by uniqueone
,

Google의 머신러닝 엔지니어링 실무 지침서입니다.

머신러닝 프로젝트 구조화에 대해 공부하다가 찾게 된 문서인데, 내용이 너무 좋아 공유 드립니다. 좀 길긴하지만, 1독 하시는걸 추천드립니다.

[https://developers.google.com/machine-learning/guides/rules-of-ml/?hl=ko](https://developers.google.com/machine-learning/guides/rules-of-ml/?hl=ko)

Posted by uniqueone
,

안녕하세요. Tensorflow KR

카카오브레인에서 자연어처리 라이브러리인 Pororo를 출시하게 되어서 글을 작성합니다. Pororo는 영어, 한국어, 중국어, 일본어 등 여러가지 언어로 30가지 이상의 자연어 처리 모델이 구현되어 있는 파이썬 라이브러리 입니다.

인공지능이나 자연어처리를 전혀 모르더라도 3~4 줄 정도의 코드만 작성하면 개체명 인식, 기계 독해, 기계 번역, 요약, 감정분류 등 다양한 태스크를 손쉽게 수행할 수 있습니다. 예를 들면 NER은 아래와 같은 코드로 수행할 수 있습니다.

pip install pororo로 설치하시면 바로 사용해보실 수 있으며 언어별 모델, 태스크는 확장해나갈 계획이니 많은 사용과 피드백 부탁드립니다.

보다 더 자세한 사항은 https://github.com/kakaobrain/pororo 에서 확인할 수 있습니다. 감사합니다.

Posted by uniqueone
,

Self-Supervised 계열의 paper를 정리하려는 목적으로 만든 repository에 꾸준히 스타가 생기고 있어서 기분이 좋아 정리를 할 겸 그림을 그려보았습니다.

혹시나, Self-Supervised 계열의 방법론에 관심이 있으셨던 분들은 참고해주시면 좋을 듯 하네요.

확실히 최근들어서는 contrastive learning 계열의 방법들이 주를 이루며 달려가고 있는 것 같습니다.

+ 최근에는 negative sample에 대한 비판적인 시각과 함께 positive sample 만을 활용하는 방법론들이 각광받고 있는 것 같습니다. 또 6개월 후의 모습은 어떠할지 벌써 기대가 되는군요.

+ Dense 라고 표기해놓은 방법들은 pixel-level 로 contrastive learning을 접근하였습니다. 최근에 흥미가 많이 가는 분야입니다. 영상의 제일 기본적인 가정인 '인접 픽셀은 유사한 성격을 지닐 것이다.' 를 잘 살리지 않았나 하는 생각이 드네요.

+ 레포는 꾸준히 업데이트를 하려고 노력하고 있습니다.

+ GPU로 몰아붙이는 학습의 새로운 매커니즘이 생길까봐 무섭지만, 늘 그렇듯 어디선가에서는 효율적인 학습의 패러다임을 제시하는 방법이 등장할 것이라 믿습니다 (SimCLR --> MoCO, ViT-->DeiT). 언뜻보니 구글과 페이스북의 구도인가요..? ㅎㅎ

레포 : https://github.com/Sungman-Cho/Awesome-Self-Supervised-Papers

Posted by uniqueone
,

안녕하세요!
NLP를 공부하시는 분들, 혹은 처음 접하는 분들께 도움이 될까 하여 올려봅니다! 🙂

빅데이터 연합동아리 투빅스에서 7주동안 텍스트 세미나를 진행했습니다.
Stanford CS224n winter2019 : Natural Language Processing with Deep Learning 커리큘럼을 따라 진행하였으며, 블로그에 세미나 내용을 정리했습니다.

유명 자연어처리 강의인 CS224n을 공부하실 때, 저희 리뷰 블로그를 참고하시고 자연어처리 기본기를 다지는 데에 도움이 되셨으면 좋겠습니다!
감사합니다.

✔️ 2020 투빅스 텍스트 세미나 velog : https://velog.io/@tobigs-text1314/series

Posted by uniqueone
,

www.facebook.com/groups/TensorFlowKR/permalink/1411658372508550

 

GPT3-small / T5-base / Bertshared-base 모델을 공개합니다.

Tensorflow와 pytorch모두에서 사용이 가능합니다. 이제 한국어도 최신 언어모델을 기반으로 연구 하실 수 있으면 좋겠습니다.
간단한 예시를 colab으로 실행시켜 볼 수 있도록 만들어 두었습니다.
아직 training중인 모델은 업데이트가 될 예정이며, T5의 경우 아직 완전히 검증하진 못하였습니다.
현재 비공개 상태의 모델들도 추후 기회가 된다면 공개하도록 하겠습니다. 다양한 NLP 연구와 모델 사용에 대한 궁금증은 메세지로 남겨주십시오.

감사합니다.

github.com/kiyoungkim1/LMkor?fbclid=IwAR2094LyHcKFdQJcv8pfUFsCbWk74SVF9BihGMQi5wiIg6ExqC1Syx6mEM8

Posted by uniqueone
,

주로 YouTube에서 'Tensorflow 2.0: Deep Learning and Artificial Intelligence'로 검색해서 강의찾음. Udacity빼고.

 

- TensorFlow 2.0 Beginner Tutorials : Aladdin Persson

youtube.com/playlist?list=PLhhyoLH6IjfxVOdVC1P1L5z5azs0XjMsb

- Intro to TensorFlow for Deep Learning : Udacity

www.udacity.com/course/intro-to-tensorflow-for-deep-learning--ud187

- TensorFlow 2.0 Complete Course - Python Neural Networks for Beginners Tutorial : freeCodeCamp.org

youtu.be/tPYj3fFJGjk

- TensorFlow 2.0 Tutorial For Beginners | TensorFlow Demo | Deep Learning & TensorFlow | Simplilearn : Simplilearn

youtu.be/QPDsEtUK_D4

- Deep Learning With Tensorflow 2.0, Keras and Python : codebasics

youtube.com/playlist?list=PLeo1K3hjS3uu7CxAacxVndI4bE_o3BDtO

- Keras with TensorFlow Course - Python Deep Learning and Neural Networks for Beginners Tutorial : freeCodeCamp.org

youtu.be/qFJeN9V1ZsI

- TensorFlow 2.0 Beginner Tutorials : Aladdin Persson

youtube.com/playlist?list=PLhhyoLH6IjfxVOdVC1P1L5z5azs0XjMsb

- TensorFlow 2.0 Tutorials for Beginners : KGP Talkie

youtube.com/playlist?list=PLc2rvfiptPSR3iwFp1VHVJFK4yAMo0wuF

- Machine Learning (많은 딥러닝 동영상들 리스트 모아놓음) : freeCodeCamp.org

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
Posted by uniqueone
,

안녕하세요 딥러닝 논문읽기 모임 입니다!

오늘 소개해 드릴 논문은 현재 많은 Image classification 분야에서 SOTA를 달성했던 논문인 Sharpness-Aware Minimization for
Efficiently Improving Generalization 이라는 논문입니다.
오늘 논문 리뷰는 펀디멘탈팀의 이재윤님께서 진행해 주셨습니다.

발표자료 :https://www.slideshare.net/taeseonryu/sharpnessaware-minimization-for-efficiently-improving-generalization
발표된 논문 리스트 : https://github.com/Lilcob/-DL_PaperReadingMeeting

overparameterized 된 모델에서 training loss는 모델의 generalization 대한 보장을 하지 않고 있는게 현실 입니다.
실제로 일반적으로 수행되는 것처럼 training loss만 최적화하면 suboptimal 모델 quality로 이어집니다.
논문에서 소개하는 Sharpness-Aware Minimization (SAM)은 다양한 global minima 중에, 제너럴한 성능을 나타내는 글로벌 미니마를 찾아 성능향상에 기여를 하여 여러 태스크에서 기존의 모델링 방법에 SAM 방법론을 적용하여 다양한 분야에서 SOTA를 달성하였습니다.

https://youtu.be/iC3Y85W5tmM

Posted by uniqueone
,

https://www.datasource.ai/en/data-science-articles/21-resources-for-learning-math-for-data-science

 

21 Resources for Learning Math for Data Science

This is probably one of the biggest worries of those starting in the area of data science, learning/refreshing mathImage by DataSource.aiLet’s be honest, most people didn’t do very well in math in school, maybe not even in college, and this is ver...

www.datasource.ai




This is probably one of the biggest worries of those starting in the area of data science, learning/refreshing math

Image by DataSource.ai

Let’s be honest, most people didn’t do very well in math in school, maybe not even in college, and this is very scary and creates a barrier for those who want to explore this discipline called data science.

A few days ago I published a post in Towards Data Science and right here on our blog called Study Plan for Learning Data Science Over the Next 12 Months, where I gave some quarterly recommendations and made an emphasis on studying mathematics and statistics for this first quarter, and from which I received many questions about exactly which materials I recommended. Well, this post answers those questions. But before that, I want to give you a context.

Leaving aside the factors or reasons that have led most people to hate math, it is a reality that we need it in data science. For me, one of the biggest shortcomings I found in mathematics was its lack of applicability in the real world, I didn’t see a reason for intermediate and advanced mathematics, such as multivariate calculus. I confess that in school and college I didn’t like them for that reason, but I always did well and got good scores and averages above the majority (especially in statistics). But I still didn’t see how I could use a derivative or a matrix in the real world. I finally ended up as a software engineer and once I entered the world of data science I was able to make the connection between mathematics, statistics, and the real world.

On the other hand, it is important to clarify that we do not need a master’s degree in pure mathematics to do data science projects. As I mentioned in previous posts there is a big debate in the community about how much math we need to do a good job as data scientists.

We could say that data science is divided into two major fields of work: research and production

By research, we mean the part of research and development, which normally takes place within a large company (usually a tech company), or which has focused on cutting-edge technological issues (such as medical research). Or it is also an area that is developed within universities. This sector has very limited job offers. 

  • The great advantage is the deep knowledge of algorithms and their implementations, as well as being a person capable of creating variations of existing algorithms, to improve them. Or even create new machine learning algorithms. 
  • The disadvantage is the unpractical nature of their work. It is a very theoretical work, in which often the only objective is to publish papers and is far from the business use cases in general. For reference on this, I recently read this post on Reddit, I recommend you read it.

By production, we refer to the practical side of this discipline, where you’ll use generally and in your day to day job libraries such as scikit-learn, Tensorflow, Keras, Pytorch, and others. These libraries operate like a black box, where you enter data, you get an output, but you don’t know in detail what happened in the process. This also has its advantages and disadvantages, but it certainly makes life much easier when putting useful models into production. What I don’t recommend is to use them blindly, where you don’t have the minimum bases of mathematics to understand a little of their fundamentals and that is the objective of this post, to guide you and recommend you some valuable resources to have the necessary bases and not to operate blindly those libraries.

So if you decide to focus on Research and Development, you are going to need mathematics and statistics in depth (very in-depth). If you are going to go for the practical part, the libraries will help you deal with most of it, under the hood. It should be noted that most job offers are in the practical side.

Well, after the previous remarks, it is time to define which are the specific topics needed to have an initial basis in mathematics for data science. 

  • Linear Algebra: This subject is important to have the fundamentals of working with data in vector and matrix form, to acquire skills to solve systems of linear algebraic equations, and to find the basic matrix decompositions and the general understanding of their applicability.
  • Calculus: Here it is important to study functional maps, limits (in case of sequences, functions of one and several variables), differentiation (from a single variable to multiple cases), integration, thus sequentially building a foundation for basic optimization. It is also important here to study gradient descent.
  • Probability theory: Here you should learn about random variables, i.e. a variable whose values are determined by a random experiment. Random variables are used as a model for the data generation processes we want to study. The properties of the data are deeply linked to the corresponding properties of the random variables, such as expected value, variance, and correlations.

Note: these subjects are much deeper than what I just mentioned, this is simply a guide of the subjects and resources recommended to approach mathematics in the field of data science.

Now that we have a better idea of the path we should take, let’s examine the recommended resources to address this topic. We will divide them into basic, intermediate, and advanced. In the advanced ones, we’ll have resources focused on deep learning

Basics: in this first section of resources we’ll recommend the mathematical basics. Mathematical thinking, algebra, and how to implement math with python.

1- Introduction to mathematical thinking

 

Price: Free

Image by Coursera

Description: Learn how to think the way mathematicians do — a powerful cognitive process developed over thousands of years.

Mathematical thinking is not the same as doing mathematics — at least not as mathematics is typically presented in our school system. School math typically focuses on learning procedures to solve highly stereotyped problems. Professional mathematicians think a certain way to solve real problems, problems that can arise from the everyday world, or from science, or from within mathematics itself. The key to success in school math is to learn to think inside-the-box. In contrast, a key feature of mathematical thinking is thinking outside-the-box — a valuable ability in today’s world. This course helps to develop that crucial way of thinking.

Link: https://www.coursera.org/learn/mathematical-thinking#

2- Mathematical Foundation for AI and Machine Learning

 

Price: $46.99 usd

Image by Packt

Description: Artificial Intelligence has gained importance in the last decade with a lot depending on the development and integration of AI in our daily lives. The progress that AI has already made is astounding with innovations like self-driving cars, medical diagnosis and even beating humans at strategy games like Go and Chess. The future for AI is extremely promising and it isn’t far from when we have our own robotic companions. This has pushed a lot of developers to start writing codes and start developing for AI and ML programs. However, learning to write algorithms for AI and ML isn’t easy and requires extensive programming and mathematical knowledge. Mathematics plays an important role as it builds the foundation for programming for these two streams. And in this course, we’ve covered exactly that. We designed a complete course to help you master the mathematical foundation required for writing programs and algorithms for AI and ML.

Link: https://www.packtpub.com/product/mathematical-foundation-for-ai-and-machine-learning-video/9781789613209

3- Math for Programmers

 

Price: $47.99

Image by Manning

Description: In Math for Programmers you’ll explore important mathematical concepts through hands-on coding. Filled with graphics and more than 300 exercises and mini-projects, this book unlocks the door to interesting–and lucrative!–careers in some of today’s hottest fields. As you tackle the basics of linear algebra, calculus, and machine learning, you’ll master the key Python libraries used to turn them into real-world software applications.

Link: https://www.manning.com/books/math-for-programmers

4- Algebra 1

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/algebra

5- Algebra 2

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/algebra2

6- Master Math by Coding in Python

 

Price: $12.99

Image by Udemy

Description: You can learn a lot of math with a bit of coding!

Many people don’t know that Python is a really powerful tool for learning math. Sure, you can use Python as a simple calculator, but did you know that Python can help you learn more advanced topics in algebra, calculus, and matrix analysis? That’s exactly what you’ll learn in this course.

This course is a perfect supplement to your school/university math course, or for your post-school return to mathematics.

Let me guess what you are thinking:

  • “But I don’t know Python!”That’s okay! This course is aimed at complete beginners; I take you through every step of the code. You don’t need to know anything about Python, although it’s useful if you already have some programming experience.
  • “But I’m not good at math!”You will be amazed at how much better you can learn math by using Python as a tool to help with your courses or your independent study. And that’s exactly the point of this course: Python programming as a tool to learn mathematics. This course is designed to be the perfect addition to any other math course or textbook that you are going through.

Link: https://www.udemy.com/course/math-with-python/

7- Introduction to Linear Models and Matrix Algebra

 

Price: Free

Image by Edx

Description: Matrix Algebra underlies many of the current tools for experimental design and the analysis of high-dimensional data. In this introductory online course in data analysis, we will use matrix algebra to represent the linear models that commonly used to model differences between experimental units. We perform statistical inference on these differences. Throughout the course we will use the R programming language to perform matrix operations.

Given the diversity in educational background of our students we have divided the series into seven parts. You can take the entire series or individual courses that interest you. If you are a statistician you should consider skipping the first two or three courses, similarly, if you are biologists you should consider skipping some of the introductory biology lectures. Note that the statistics and programming aspects of the class ramp up in difficulty relatively quickly across the first three courses. You will need to know some basic stats for this course. By the third course will be teaching advanced statistical concepts such as hierarchical models and by the fourth advanced software engineering skills, such as parallel computing and reproducible research concepts.

Link: https://www.edx.org/course/introduction-to-linear-models-and-matrix-algebra

8- Applying Math with Python

 

Price: $20.99

Image by Packt

Description: Python, one of the world’s most popular programming languages, has a number of powerful packages to help you tackle complex mathematical problems in a simple and efficient way. These core capabilities help programmers pave the way for building exciting applications in various domains, such as machine learning and data science, using knowledge in the computational mathematics domain.

The book teaches you how to solve problems faced in a wide variety of mathematical fields, including calculus, probability, statistics and data science, graph theory, optimization, and geometry. You’ll start by developing core skills and learning about packages covered in Python’s scientific stack, including NumPy, SciPy, and Matplotlib. As you advance, you’ll get to grips with more advanced topics of calculus, probability, and networks (graph theory). After you gain a solid understanding of these topics, you’ll discover Python’s applications in data science and statistics, forecasting, geometry, and optimization. The final chapters will take you through a collection of miscellaneous problems, including working with specific data formats and accelerating code.

By the end of this book, you’ll have an arsenal of practical coding solutions that can be used and modified to solve a wide range of practical problems in computational mathematics and data science.

Link: https://www.packtpub.com/product/applying-math-with-python/9781838989750




Intermediate: in this second section we will recommend resources focused on calculation and probability.

9- Calculus 1

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/calculus-1

10- Calculus 2

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/calculus-2

11- Multivariable calculus

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/multivariable-calculus

12- Mathematics for Data Science Specialization

 

Price: Free

Image by Coursera

Description: Behind numerous standard models and constructions in Data Science there is mathematics that makes things work. It is important to understand it to be successful in Data Science. In this specialisation we will cover wide range of mathematical tools and see how they arise in Data Science. We will cover such crucial fields as Discrete Mathematics, Calculus, Linear Algebra and Probability. To make your experience more practical we accompany mathematics with examples and problems arising in Data Science and show how to solve them in Python.

Each course of the specialisation ends with a project that gives an opportunity to see how the material of the course is used in Data Science. Each project is directed at solving practical problem in Data Science. In particular, in your projects you will analyse social graphs, predict estate prices and uncover hidden relations in the data.

Link: https://www.coursera.org/specializations/mathematics-for-data-science

13- Practical Discrete Mathematics

 

Price: $24.99

Image by Packt

Description: Discrete mathematics is a field of math that deals with studying finite and distinct elements. The theories and principles of discrete math are widely used in solving complexities and building algorithms in computer science and computing data in data science. It helps you to understand algorithms, binary, and general mathematics that is commonly used in data-driven tasks.

Learn Discrete Mathematics is a comprehensive introduction for those who are new to the mathematics of countable objects. This book will help you get up-to-speed with implementing discrete math principles to take your programming skills to another level. You’ll learn the discrete math language and methods crucial to studying and describing objects and functions in branches of computer science and machine learning. Complete with real-world examples, the book covers the internal workings of memory and CPUs, analyzes data for useful patterns, and shows you how to solve problems in network routing, encryption, and data science.

By the end of this book, you’ll have a deeper understanding of discrete mathematics and its applications in computer science, and get ready to work on real-world algorithm development and machine learning.

Link: https://www.packtpub.com/product/practical-discrete-mathematics/9781838983147

14- Math for Data Science and Machine Learning: University Level

 

Price: $12.99

Image by Udemy

Description: In this course we will learn math for data science and machine learning. We will also discuss the importance of Math for data science and machine learning in practical word. Moreover, Math for data science and machine learning course is bundle of two courses of linear algebra and probability and statistics. So, students will learn complete contents of probability and statistics and linear algebra. It is not like that you will not complete all the contents in this 7 hours videos course. This is a beautiful course and I have designed this course according to the need of the students.

Linear algebra and probability and statistics is usually offered for the students of data science, machine learning, python and IT students. So, that’s why I have prepared this dual course for different sciences.

I have taught this course multiple times on my universities classes. It is offered usually in two different modes like, it is offered as linear algebra for 100 marks paper and probability and statistics as another 100 marks paper for two different or in a same semesters. I usually focus on the method and examples while teaching this course. Examples clear the concepts of the students in a variety of way like, they can understand the main idea that instructor want to deliver if they feel typical the method of the subject or topics. So, focusing on example makes the course easy and understandable for the students.

Link: https://www.udemy.com/course/master-linear-algebra-and-probability-2-in-1-bundle/

15- Data Science Math Skills

 

Price: Free

Image by Coursera

Description: Data science courses contain math — no avoiding that! This course is designed to teach learners the basic math you will need in order to be successful in almost any data science math course and was created for learners who have basic math skills but may not have taken algebra or pre-calculus. Data Science Math Skills introduces the core math that data science is built upon, with no extra complexity, introducing unfamiliar ideas and math symbols one-at-a-time.

Learners who complete this course will master the vocabulary, notation, concepts, and algebra rules that all data scientists must know before moving on to more advanced material.

Link: https://www.coursera.org/learn/datasciencemathskills




Advanced: in this last section we will focus on the statistical part (probability theory) and the application of mathematics to deep learning algorithms.

16- Statistics and probability

 

Price: Free

Image by Khanacademy

Link: https://www.khanacademy.org/math/statistics-probability

17- Intro to Inferential Statistics

 

Price: Free

Image by Udacity

Description: Inferential statistics allows us to draw conclusions from data that might not be immediately obvious. This course focuses on enhancing your ability to develop hypotheses and use common tests such as t-tests, ANOVA tests, and regression to validate your claims.

Link: https://www.udacity.com/course/intro-to-inferential-statistics--ud201

18- Statistical Methods and Applied Mathematics in Data Science

 

Price: $124.99

Image by Packt

Description: Machine learning and data analysis are the center of attraction for many engineers and scientists. The reason is quite obvious: its vast application in numerous fields and booming career options. And Python is one of the leading open source platforms for data science and numerical computing. IPython, and its associated Jupyter Notebook, provide Python with efficient interfaces to for data analysis and interactive visualization, and they constitute an ideal gateway to the platform. If you are among those seeking to enhance their capabilities in machine learning, then this course is the right choice.

Statistical Methods and Applied Mathematics in Data Science provides many easy-to-follow, ready-to-use, and focused recipes for data analysis and scientific computing. This course tackles data science, statistics, machine learning, signal and image processing, dynamical systems, and pure and applied mathematics. You will apply state-of-the-art methods to various real-world examples, illustrating topics in applied mathematics, scientific modeling, and machine learning. In short, you will be well versed with the standard methods in data science and mathematical modeling.

Link: https://www.packtpub.com/product/statistical-methods-and-applied-mathematics-in-data-science-video/9781789539219

19- Exploring Math for Programmers and Data Scientists

 

Price: Free

Image by Manning

Description: Exploring Math for Programmers and Data Scientists showcases chapters from three Manning books, chosen by author and master-of-math Paul Orland. You’ll start with a look at the nearest neighbor search problem, common with multidimensional data, and walk through a real-world solution for tackling it. Next, you’ll delve into a set of methods and techniques integral to Principal Component Analysis (PCA), an underlying technique in Latent Semantic Analysis (LSA) for document retrieval. In the last chapter, you’ll work with digital audio data, using mathematical functions in different and interesting ways. Begin sharpening your competitive edge with the fun and fascinating math in this (free!) practical guide!

Link: https://www.manning.com/books/exploring-math-for-programmers-and-data-scientists

20- Hands-On Mathematics for Deep Learning

 

Price: $27.99

Image by Packt

Description: Most programmers and data scientists struggle with mathematics, having either overlooked or forgotten core mathematical concepts. This book uses Python libraries to help you understand the math required to build deep learning (DL) models.

You’ll begin by learning about core mathematical and modern computational techniques used to design and implement DL algorithms. This book will cover essential topics, such as linear algebra, eigenvalues and eigenvectors, the singular value decomposition concept, and gradient algorithms, to help you understand how to train deep neural networks. Later chapters focus on important neural networks, such as the linear neural network and multilayer perceptrons, with a primary focus on helping you learn how each model works. As you advance, you will delve into the math used for regularization, multi-layered DL, forward propagation, optimization, and backpropagation techniques to understand what it takes to build full-fledged DL models. Finally, you’ll explore CNN, recurrent neural network (RNN), and GAN models and their application.

By the end of this book, you’ll have built a strong foundation in neural networks and DL mathematical concepts, which will help you to confidently research and build custom models in DL.

Link: https://www.packtpub.com/product/hands-on-mathematics-for-deep-learning/9781838647292

21- Math and Architectures of Deep Learning

 

Price: $39.99

Image by Manning

Description: Math and Architectures of Deep Learning sets out the foundations of DL in a way that’s both useful and accessible to working practitioners. Each chapter explores a new fundamental DL concept or architectural pattern, explaining the underpinning mathematics and demonstrating how they work in practice with well-annotated Python code. You’ll start with a primer of basic algebra, calculus, and statistics, working your way up to state-of-the-art DL paradigms taken from the latest research. By the time you’re done, you’ll have a combined theoretical insight and practical skills to identify and implement DL architecture for almost any real-world challenge.

Link: https://www.manning.com/books/math-and-architectures-of-deep-learning

Conclusion

This is an extensive recommendation on resources for learning mathematics for data science, following the previous post about the path to follow in this year 2021 to learn data science

When we have limited time for study, we should select those that we feel best and those that fit our style. For example, you might prefer videos about books, so go ahead and choose what suits you best. This material is sufficient whether you want to take a brief look at the mathematics, or if you want to go deeper into it. I hope you find it useful.

If you have other recommendations for courses, books or videos, please leave them in the comments so that we can all create links of interest.

Note: we are building a private community in Slack of data scientist, if you want to join us you can register here: https://www.datasource.ai/en#slack

I hope you enjoyed this reading! you can follow me on twitter or linkedin

Thanks for reading!



Posted by uniqueone
,

https://www.facebook.com/groups/KaggleKoreaOpenGroup/permalink/828642621201172/

김선우

안녕하세요. 평소 머신러닝, 딥러닝에 관심이 많았는데 캐글은 이번에 처음 시작하게 되어 질문 드립니다. 1. 연구에서도 그렇지만 캐글에서는 특히 모델이 좋은 성능을 내도록 최적의 방법을

www.facebook.com

안녕하세요. 평소 머신러닝, 딥러닝에 관심이 많았는데 캐글은 이번에 처음 시작하게 되어 질문 드립니다.
1. 연구에서도 그렇지만 캐글에서는 특히 모델이 좋은 성능을 내도록 최적의 방법을 사용하는 것이 중요한 것 같습니다. 그러기 위해서는 최신 연구들도 참고할 필요가 있는데 최근에 논문들이 쏟아져 나오고 있는 만큼 이를 따라가기가 어렵다고 느꼈습니다. 캐글 노트북과 discussion들을 참고하는 것도 한 방법인 것 같은데 이 외에 최적의, 최신의 방법론들을 빠르게 습득하고 따라가기 위한 좋은 방법들이 있을까요?
2. 캐글 노트북은 9시간 후에 끊긴다는 단점이 있고 매번 커밋해야지만 학습 과정이 저장되는 것으로 알고 있는데 맞나요? 이러한 점 때문에 대회에서 캐글 노트북을 사용하기 어려운 경우가 있을 것 같은데 원래 다른 곳(코랩, 개인 하드웨어 등)에서 코드를 돌려보고 결과물 또는 코드를(대회에 따라) 제출하는 것이 일반적인가요?
감사합니다.

1. 커뮤니티 가입하셔서 올라오는 글들 보시거나,
탑티어학회 논문들 챙겨 보시면됩니다.

2. 네, 보통 다른 머신에서 학습하고 제출 또는
인퍼런스만 노트북에서 합니다.

안녕하세요. 캐글 입문을 축하드립니다.!
한국의 캐글 그랜드마스터 수가 ? 나라보다는 많아야 한다고 믿는 Limerobot입니다. 저는 지금까지 캐글에서 금메달 8개를 획득했었는데요, 이 경험을 바탕으로 저가 애용하는 방법을 공유 드리겠습니다.

1.
우선 캐글은 극히 실용성을 추구하기에 실제로 성능이 좋고 사용하기 쉬운 모델이나 라이브러리만 활용합니다. (왜냐면 이 외에도 할 일이 엄청 많거든요) 그래서 참여할 대회와 비슷한 종료된 대회의 상위 솔루션을 참고하시면 도움이 됩니다. 보통 좋은 것이 우승하거든요.

아래 링크는 대회 중에도 제가 계속 참고하는 곳입니다.

# 이미지 분야
최신 백본망은 아래 링크 참고
https://paperswithcode.com/sota/image-classification-on-imagenet

위의 백본망 중 아래 파이썬 패키지에서 활용 가능한 것 추천
https://github.com/qubvel/segmentation_models.pytorch

# 텍스트 분야
최신 백본망은 아래 링크를 참고
https://gluebenchmark.com/leaderboard

위의 백본망 중 아래 파이썬 패키지에서 활용 가능한 것 추천
https://github.com/huggingface/transformers

# 그 외 분야
해당 대회의 주최측에서 제공하는 정보나 대회 참여자들이 공유하는 discussion, notebooks 탭을 참고 또는 이전에 수행된 비슷한 대회를 참고하시는 게 좋습니다.










Posted by uniqueone
,

"RepVGG: Making VGG-style ConvNets Great Again" 논문 소개

안녕하세요, Cognex 이호성입니다. 재미있는 논문이 3일전에 공개되어서 소개드립니다.
딥러닝을 이용한 Computer Vision을 처음 공부하면 대표적인 architecture들 부터 배우게 됩니다. 보통 LeNet, AlexNet을 거쳐서 VGG로 넘어가죠? VGG는 다들 한번쯤은 들어보셨을겁니다.
CNN architecture의 발전 과정은 제가 외부에서 발표했던 자료를 참고하시면 도움이 되실 것 같아서 먼저 소개드립니다.
- CNN architecture 톺아보기: https://www.slideshare.net/HoseongLee6/cnn-architecture-a-to-z

CNN architecture는 ResNet의 대 유행을 지나 AutoML을 이용한 Neural Architecture Search(NAS) 계열의 EfficientNet, RegNet이 좋은 성능을 내고 있었는데 갑자기 예전의 추억을 떠오르게 하는 VGG 연구가 제안되었습니다.
논문 링크: https://arxiv.org/abs/2101.03697v1

아이디어는 단순합니다. 그림과 같이 VGG에 기존 ResNet (그림 A)에서 사용하던 residual branch를 추가하여 Single-path에서 Multi-branch로 VGG를 학습시키게 됩니다. (그림 B)
정확도가 높아지면 대체로 연산 처리 속도는 느려지기 마련인데, 저자들은 딥러닝 연산에 사용되는 NVIDIA cuDNN, Intel MKL 등의 library가 3x3 conv에 굉장히 최적화되어있다는 점에 주목하여 그림 C와 같이 3x3 연산만 거쳐서 inference를 합니다.
여기서 발생하는 Train - Test 의 모델의 불일치를 완화시키기 위해 structural re-parameterization 이라는 기법을 적용합니다. 자세한 내용은 첨부드린 그림과 논문을 참고하시면 좋을 것 같습니다.
실험 결과 ResNet 계열보다 더 적은 Parameter 수로 더 높은 정확도와 더 빠른 처리속도를 달성할 수 있었고, SOTA 방법론인 EfficientNet, RegNet과도 견줄만한 accuracy-speed trade-off 그래프를 얻을 수 있었다고 합니다.
7년전 제안된 VGG가 2021년에 다시 나타날 줄은 상상도 못했는데 정말 세상엔 재미있는 연구들이 많은 것 같네요! 읽어주셔서 감사합니다.

Posted by uniqueone
,

안녕하세요. Tensorflow KR. 자연어처리 리뷰모임 [집현전]의 운영자 고현웅이라고 합니다. 이번 12월 한 달 동안 집현전에서 만든 자연어처리 논문 리뷰 비디오 8편 (초급반 4편, 중급반 4편) 을 공유하기 위해 글을 작성하였습니다. 관심 있으신 분들은 아래 설명과 영상을 참고하시면 좋을 것 같습니다!


[집현전 Links]
집현전 Github : https://github.com/jiphyeonjeon/nlp-review
집현전 초급반 Youtube : https://www.youtube.com/playlist?list=PLsXisDblbLJ_UaUPi0iThUQKgSbhIk5ho
집현전 중급반 Youtube : https://www.youtube.com/playlist?list=PLsXisDblbLJ8msozFyA8o2zf3zfbORHWd

[초급반 12월 발표]
1. 송석리 - Long Short Term Memory
https://youtu.be/HHKSCkPEQfw

자연어처리에 대표적으로 사용되는 시퀀스 처리 모델인 LSTM에 대해 리뷰합니다. LSTM을 알아보기 이전에 RNN에 대해 알아보고, LSTM의 각종 게이트의 역할, 그리고 실험 결과를 깔끔한 일러스트와 함께 설명합니다.
2. 이영빈 - Word2Vec
https://youtu.be/o-2kfiZP7Z8

워드 임베딩 모델인 Word2Vec에 대해 논의합니다. 특히 Skip-gram (주변 단어 예측) 과 CBOW (중심단어 예측) 등의 방식에 특성에 대해 논의하고 Word Analogy 등의 태스크에서 성능을 비교하였습니다.
3. 이기창 - FastText
https://youtu.be/7pDB9zqwxhs

Word2Vec을 개선한 FastText 임베딩에 대해 논의합니다. OOV 문제를 완화하기 위해 도입된 서브워드 임베딩의 개념에 대해 이야기하고 실험 성능을 비교합니다. 특히 Semantic 영역의 성능과 Syntactic 영역의 성능을 비교하여 평가합니다.
4. 허치영 - Neural Probabilistic Language Model (NPLM)
https://youtu.be/EWMNCTpfsLI

뉴럴 기반의 언어모델의 시초 격인 NPLM에 대해 리뷰합니다. NPLM의 학습 오브젝티브를 자세한 수식으로 알아보고 NPLM이 Long term dependency에 효과적인 이유에 대해 논의하고 실험 성능을 비교합니다.

[중급반 12월 발표]
1. 진명훈 - If Beam Search is the Answer, What was the Question?
https://youtu.be/KJClfF_nJj0

디코딩 전략 중 한가지인 빔서치에 대해 논의합니다. 빔 사이즈를 최대로 높히는 것과 동일한 MAP 생성이 실제로는 잘 안되는 문제를 지적하며 빔서치가 좋은 문장을 생성해내는 이유를 인지심리학적인 요인을 통해 분석합니다. 추가로 UID 디코딩이라는 새로운 기법에 대해서도 논의합니다.
2. 박동주 - SSMBA : Self Supervised Manifold Based Data Augmentation for Improving Out of Domain Robustness
https://youtu.be/1IwHZ_4uPK0

Vicinal Risk Minimization의 일환으로 BERT의 MLM을 활용한 Data Augmentation 기법에 대해 논의합니다. Machine Translation, NLI, Sentiment Analysis 등의 Downstram Task에서 실험을 진행하였으며, 특히 In domain과 Out of domain 세팅에서의 성능을 비교합니다.
3. 고현웅 - Machine Translation Survey
1편 : https://youtu.be/KQfvEg-fGMw
2편 : https://youtu.be/18iH6VX-IU4

총 4시간에 걸쳐서 Seq2Seq부터 최신 연구(mBART, M2M-100, M2 등...)까지 23편의 논문을 참고하여 신경망 기계 번역의 역사와 최신 연구 동향에 대해 논의합니다. Background, Dataset, Efficiency, Training, Architecture 등에 대해 논의를 진행하고 성능, 서비스, 앞으로의 방향성에 대해 디스커션을 진행합니다.

Posted by uniqueone
,

텐서플로우 입문 강의를 두 개 제작하여 머신러닝 야학에 공개했는데요.
열심히 만들었는데... TensorFlow KR 그룹에 공유해봅니다. :)

표 형태의 데이터로 지도학습의 회귀와 분류를 경험하는 강의입니다.
텐서플로우를 이용하여 딥러닝 모델을 구현하고, 딥러닝의 기본적인 개념을 배웁니다.
Tensorflow 101: https://www.youtube.com/watch?v=auCw6qikSYs&list=PLl1irxoYh2wyLwJutUZx5Q_QEEDZoXBnz

이미지로 분류 모델인 CNN을 만들어보는 강의입니다.
텐서플로우를 이용하여 CNN 모델을 구현하고, CNN의 기본적인 개념을 배웁니다.
Tensorflow 102: https://www.youtube.com/watch?v=MMEoEJIXd7E&list=PLl1irxoYh2wzOOU9hvJqMYc215wAlxrpp

코딩 없이 인공지능을 경험해보고 싶은 분은 머신러닝 1 수업을 추천합니다.
https://opentutorials.org/course/4548
이 수업은 13개의 동영상으로 이루어진 2시간 분량의 수업입니다.

자세한 내용과 참가신청은 머신러닝 야학 홈페이지를 참고해주세요.
https://ml.yah.ac/

Posted by uniqueone
,

새해부터 데이터셋 Bomb이 터졌네요. Eleuther AI 에서 공개적으로 수집 및 훈련에 사용이 가능한 영어 텍스트 825GB 코퍼스 Pile을 공개하였습니다.

저자진은 다양한 토픽으로 구성된 코퍼스로 언어 모델을 학습시켰을 때, 다운스트림 태스크에서 일반화가 잘 된다는 최근 연구 결과들을 따라 다양한 주제의 코퍼스를 모으기 위해 특히 신경을 많이 썼다고 합니다.

이외에도 코퍼스에서 등장하는 인종별, 성별 바이어스 등에 대한 분석, 경멸적 표현들에 대한 분석도 굉장히 Time-consuming 하지만 수행을 했다고 하네요 👍

또한 GPT-2 사이즈의 모델을 자신들이 공개한 Pile에 대해 학습시켰을 때, CC-100 (영어)에 대해 학습된 모델보다 여러 도메인에서 좋은 성능을 보이는 것까지 실험을 통해 확인하였습니다.

커뮤니티에서는 GPT-X의 democratize를 위한 첫 번째 움직임이라며, 굉장히 많은 호응을 얻고 있습니다 🤭

link: https://pile.eleuther.ai/

Posted by uniqueone
,

안녕하세요. 제 깃헙에 아래 첨부한 사진에 나오는 모델들을 개인 공부 차원으로 구현해 보았으니 필요하신 분들은 편하게 가져가세요~~~~

PS : Multi-Head Attention 기능도 포함 되어 있습니다.!
https://github.com/jk96491/Advanced_Models

jk96491/Advanced_Models

여러가지 유명한 신경망 모델들을 제공합니다. (DCGAN, VAE, Resnet 등등). Contribute to jk96491/Advanced_Models development by creating an account on GitHub.

github.com

러가지 유명한 신경망 모델들을 제공합니다.
제공 모델들 : DCGAN, CGAN, SA-GAN ,GAN, Resnet, VAE, Multi-Head Attention, GPT-2

Posted by uniqueone
,

안녕하세요 Tensorflow KR.

이번에 제가 하고있는 논문스터디에서 기계번역 서베이를 준비하면서 자연어처리 백그라운드(Seq2Seq 부터 BERT까지)에 대한 내용을 함께 준비 해봤는데요. NLP 전반적으로 폭넓게 다루고 있는 영상이라 자연어처리 공부를 시작한지 얼마 안되신 분들이라면 도움 되실분들이 있을 것 같아서 여기에도 업로드 해봅니다.

영상에서 "왜?" 라는 질문을 많이 포함하고 있어서(왜 multihead? 왜 layer norm? 왜 pos embeddimg? 등..) 어텐션이나, Tranaformer 등에 대해 대략적으로만 알고 뭔가 와닿지는 않는 그런 분들이 들으시면 아마 도움이 더 많이 되시지 않을까 생각합니다.(백그라운드만 1시간 발표라, 배보다 배꼽이 더 커져서 결국 발표를 두번이나 했네요ㅋㅋㅋㅋㅠ)

참고로 집현전 추가 모집 문의가 정말 많이 왔는데요. 현재로서는 일정이나 깃헙, 영상등 관리하는게 빠듯해서 당장은 어려울 것 같구요. 시즌1을 성공적으로 마치게 되면 (모든분들이 한번씩 발표하고나서) 새 인원모집에 대해 다시 공지를 드리려고 합니다. 집현전 영상의 경우는 매달 저희가 읽은 논문과 자료들을 TFKR에 제가 정리해서 업로드 해드리겠습니다! 감사합니다~https://m.youtube.com/watch?v=KQfvEg-fGMw&feature=youtu.be

Posted by uniqueone
,

대규모 한국어 텍스트 데이터로 학습한 pre-training 언어 모델을 공개합니다!!​

블로그, 리뷰, 댓글과 같이 사람들이 직접 사용하는 텍스트를 수집하고 전처리하여 얻은 70Gb의 데이터로 Bert, Albert, Electra의 base 모델을 학습하였습니다. Pre-release버전을 공개하며 내년 초 더 많은 모델을 공개할 예정입니다.​
Huggingface를 통해 사용할 수 있으며, fine-tuning시 기존에 공개된 모델에 버금가거나 더 나은 성능을 보여주고 있습니다.​

이번 모델은 제가 프로젝트를 진행하다 공개하게 되었습니다. ​
텍스트, 이미지, 음성등의 인공지능을 활용해 다양한 비즈니스모델을 단기간 테스트하고 평가하며 좋은 서비스를 찾아가는 조직을 구성하고 있으니 관심 있으신 분들은 연락주십시오.​

자세한 사항은 아래 Github을 참고해주세요. 감사합니다!
https://github.com/kiyoungkim1/LM-kor

Posted by uniqueone
,

안녕하세요, TensorFlow KR 여러분!

생성 모델(Generative Model)은 실제로는 존재하지 않지만, 있을법한 데이터를 만들 수 있는 모델을 의미합니다. 오늘은 현대 딥러닝 기반의 생성 모델에 큰 영향을 끼친 논문인 GAN(NIPS 2014)을 소개합니다. GAN은 특히 이미지 도메인에서의 많은 발전이 이루어져 특정한 도메인의 이미지(사람의 얼굴, 동물의 형체 등)를 생성하거나 수정하는 작업을 수월하게 만들어주고 있습니다.

사실 GAN은 나온 지 오래되어 너무 잘 알려진 논문이며 아이디어는 참 직관적이지만, 증명적인 내용은 딥러닝 입문자가 처음 읽기에는 난이도가 높다고 생각합니다. 고등학교 때 문과였던 저를 포함해 많은 개발자분에게는 논문에 적힌 수식을 이해하기가 어렵고, 고등학교 때 어렴풋이 배웠던 확률 분포 개념부터 헷갈려 헤맬 수 있습니다.

본 영상에서는 GAN 아이디어 및 수식에 대한 내용 요약, PyTorch로 MNIST 이미지 생성해보는 실습, NIPS 2014에 업로드된 GAN 논문 원서를 함께 읽어보는 내용을 한꺼번에 포함하고 있습니다. 생성 모델 쪽을 처음 공부하시는 분들에게 도움이 되었으면 좋겠습니다.

영상 링크:

Posted by uniqueone
,

안녕하세요.

한국어 문장 분리기로 많은 사랑을 받고 있는 Korean Sentence Splitter(KSS)의 새로운 소식을 어디다 남길까 고민 하다가 이 그룹에서 가장 많은 분들이 사용하고 계셔서 여기에 남겨봅니다 ^^

작년 즈음에 카카오에서 사용해오던 한국어 문장 분리기에 착안해 C++로 한국어 문장 분리기를 새롭게 만들었고, 오픈소스로 공개하여 좋은 반응을 얻은 바 있습니다. 그러나 지속적인 문의나 요청 사항에는 거의 대응을 하지 못했고, 결정적으로 저 또한 전혀 다른 프로젝트를 맡게 되어 더 이상 유지보수를 할 수가 없었는데요. 특히 C++로 구현하다 보니 빌드 문의가 정말 많았고, 원래는 Windows, Mac, Linux 각 OS별로 빌드하여 바이너리를 업로드하여 배포하도록 권장하고 있는데, 상용 프로젝트도 아니고 그렇게 까지 관리할 수는 없었습니다.

결정적으로 문장 분리 자체가 엄청난 고성능을 요구해서 꼭 C++로 작성해야 하는건 아니었기에 빌드의 번거로움과 개발 유지보수를 감안하면 이제 다른 언어로 바꾸는게 좋겠다고 생각하던 차, 마침 고현웅님께서 파이썬으로 모두 포팅해주셨고, 또 꾸준히 개선해 나가는 모습을 보면서 이제는 프로젝트를 넘겨줄때가 됐음을 깨달았습니다.

제가 올 초에 1.3.1까지 올렸었고, 오늘부터 고현웅님이 만드신 파이썬 포팅 버전으로 2.0.0이 시작됩니다. 그동안 저에게 들어와있던, 제가 처리하지 못했던 모든 이슈와 PR이 반영된 최종 개선 버전이고, 아마 앞으로도 잘 개선해주시리라 기대가 큽니다.

팩키지 설치는 기존과 동일하게 `pip install kss`로 가능하며, 버그나 개선과 관련한 이슈는 새로운 레포인 https://github.com/hyunwoongko/kss 에 올려주시면 됩니다.

앞으로도 많은 응원 부탁드립니다.
감사합니다.

Posted by uniqueone
,

Lectures for UC Berkeley CS 285: Deep Reinforcement Learning
Playlist : https://www.youtube.com/playlist?list=PL_iWQOsE6TfURIIhCrlt-wj9ByIVpbfGc
H / T : Sergey Levine
#ArtificialIntelligence #DeepLearning #ReinforcementLearning

Posted by uniqueone
,

Deep Learning in Life Sciences
by Massachusetts Institute of Technology (MIT)

Course Site: https://mit6874.github.io/

Lecture Videos: https://youtube.com/playlist?list=PLypiXJdtIca5ElZMWHl4HMeyle2AzUgVB

We will explore both conventional and deep learning approaches to key problems in the life sciences, comparing and contrasting their power and limits. Our aim is to enable you to evaluate a wide variety of solutions to key problems you will face in this rapidly developing field, and enable you to execute on new enabling solutions that can have large impact.
As part of the subject you will become an expert in using modern cloud resources to implement your solutions to challenging problems, first in problem sets that span a carefully chosen set of tasks, and then in an independent project.
You will be programming using Python 3 and TensorFlow 2 in Jupyter Notebooks on the Google Cloud, a nod to the importance of carefully documenting your work so it can be precisely reproduced by others.

#artificialintelligence #deeplearning #tensorflow #python #biology #lifescience

Posted by uniqueone
,

안녕하세요. Tensorflow KR.

얼마전에 SKT에서 한국어 Seq2Seq 사전학습 모델인 KoBart를 공개하였는데요. 이를 git clone 없이 편리하게 사용할 수 있도록 huggingface transformers에 포팅하게 되어서 글을 작성합니다. `pip install kobart-transformers`로 쉽게 설치할 수 있으며, 기타 사용법은 아래 레포에서 확인할 수 있습니다. 감사합니다.


github.com/hyunwoongko/kobart-transformers?fbclid=IwAR1SRkBBe-SKGipBrjJBqMIUs8s7YKaOe8Q8eA3Q8yzJPFwbjokbOlbBiHs

Posted by uniqueone
,

안녕하세요 딥러닝 논문읽기 모임 입니다!

논문 발표자료 : https://www2.slideshare.net/taeseonryu/explaining-knowledge-distillation
지금까지 업데이트 된 논문 리뷰 영상 : https://github.com/Lilcob/-DL_PaperReadingMeeting
딥논읽 채널 : https://www.youtube.com/channel/UCDULrK2OJsiDhFroa2Aj_LQ?view_as=subscriber

오늘 업데이트된 따끈따근한 논문 리뷰 영상은
'Explaining Knowledge Distillation by Quantifying the Knowledge'
라는 제목의 논문입니다 해당논문은 2020 cvpr에서 프로시딩된 논문입니다!
해당논문은 DNN의 인풋 이후 메인 태스크와 관련이 있는 이미지와 관련이 없는 이미지의 feature를 찾아내어 시각적으로 정량화하고 분석하여 knowledge distillation을 성공적으로 수행하는방법을 제시합니다! 또한 해당논문은 세가지의 가설을 제시하여 논문의 메소드에대한 설명을 보다 탄탄하게 뒷받침 하는대요,
펀디멘탈팀의 김동희님이 자세한 논문 리뷰를 도와주셨습니다!
오늘도 많은 관심 감사드립니다!
https://youtu.be/3q3iHhgKOY0

Posted by uniqueone
,

6 free, high-quality Udemy courses for you today:

Python
https://www.udemy.com/course/python-comprehensive-bootcamp-beginner-to-professional/?ranMID=39197&ranEAID=WIRXHBzeZwo&ranSiteID=WIRXHBzeZwo-NFnZtLOHl0lU6RYjRtpMdA&LSNPUBID=WIRXHBzeZwo&utm_source=aff-campaign&utm_medium=udemyads&couponCode=25D2E97D660C1C362067

Deep Learning - Python & R
https://www.udemy.com/course/deep-learning-with-keras-and-tensorflow-in-python-and-r/?ranMID=39197&ranEAID=WIRXHBzeZwo&ranSiteID=WIRXHBzeZwo-NFnZtLOHl0lU6RYjRtpMdA&LSNPUBID=WIRXHBzeZwo&utm_source=aff-campaign&utm_medium=udemyads&couponCode=DECOUP20

Data viz in Excel
https://www.udemy.com/course/data-visualization-in-excel-for-business-professionals/?ranMID=39197&ranEAID=WIRXHBzeZwo&ranSiteID=WIRXHBzeZwo-NFnZtLOHl0lU6RYjRtpMdA&LSNPUBID=WIRXHBzeZwo&utm_source=aff-campaign&utm_medium=udemyads&couponCode=DECOUP20

Statistics/Probability
https://www.udemy.com/course/probability-and-statistics-1-the-complete-guide/?ranMID=39197&ranEAID=WIRXHBzeZwo&ranSiteID=WIRXHBzeZwo-NFnZtLOHl0lU6RYjRtpMdA&LSNPUBID=WIRXHBzeZwo&utm_source=aff-campaign&utm_medium=udemyads&couponCode=6F666FD4D86872974BA9

R
https://www.udemy.com/course/data-manipulation-with-dplyr-in-r/?ranMID=39197&ranEAID=WIRXHBzeZwo&ranSiteID=WIRXHBzeZwo-NFnZtLOHl0lU6RYjRtpMdA&LSNPUBID=WIRXHBzeZwo&utm_source=aff-campaign&utm_medium=udemyads&couponCode=MATURITY

mySQL
https://www.udemy.com/course/the-complete-sql-course-2020-become-a-mysql-master/?couponCode=SQLCOURSE29

Posted by uniqueone
,

Cheat sheet for python Neural network Machine Learning Data Science for newbies ALL PDF.https://bit.ly/394jjDH

Posted by uniqueone
,

Top 13 YouTube Channels to Learn Data Science


Get started

Open in app



497K Followers

·

About

Follow

Responses

To respond to this story,
get the free Medium app.


Open in app

There are currently no responses for this story.

Be the first to respond.

You have 2 free member-only stories left this month. 

Sign up for Medium and get an extra one

Top 13 YouTube Channels to Learn Data Science

Informative channels on YouTube to gain access to tons of informative videos on Data Science



Claire D. Costa

Nov 9·12 min read





Photo by freestocks.org from Pexels

Data science is the discipline of making data useful.

YouTube doesn’t need much of an introduction, and we’re sure you all know how much of a popular platform it is among people of all age groups. YouTube is not just a vast repository of entertainment, but it is also an equally important source of education. Undeniably the best thing about learning from YouTube is the lack of any fee or charges for getting access to that vast repertoire of educational videos.
Data Science and many other domains like it can seem daunting at first, but with YouTube, you can get easy access to a plethora of educational and instructional videos on a wealth of topics, including Data Science. The goal behind this write-up is to introduce you to a range of informative channels on YouTube that demystify the complex concepts of Data Science so that you can learn at your own pace. But before digging into this, check out
Some Interesting Articles for Data Science —

Data Science Books You Must Read in 2020

Have a look, why you should read them?

towardsdatascience.com

Best Data Science Blogs to Follow in 2020

Most trusted and reputed sources to update yourself with the latest happenings in the Data Science world.

towardsdatascience.com

10 Popular Data Science Resources on Github

Some of the top GitHub repositories that will teach you all about Data Science.

towardsdatascience.com

Top Data Science YouTube Channels

The domain of Data Science brings with itself a variety of scientific tools, processes, algorithms, and knowledge extraction systems from structured and unstructured data alike, for identifying meaningful patterns in it.

Everybody loves YouTube, right? But wouldn’t it be fun to gain access to literally thousands of informative videos on Data Science, covering not just the basics but also the latest happenings in the domain?
In the age of Massive Online Open Courses(MOOCs), YouTube serves as a powerful platform to find answers to your questions, which otherwise would not have made it to the course video. All that while helping you save some money.
We’re guessing by now you’re pretty hyped to know about the channels, fair enough, let’s not waste any more time and move on to the list of YouTube channels focused on Data Science.

1. 3Blue1Brown





Source: 3Blue1Brown

Since: Mar 4, 2015

Creator: Grant Sanderson

Views: 165,573,366

Subscribers: 3.22M

Youtube Link: https://www.youtube.com/channel/UCYO_jab_esuFRV4b17AJtAw

3Blue1Brown is a fairly enjoyable channel created by Grant Sanderson in March 2015 that primarily focuses on teaching math in an entertaining way.
You might be wondering why we included this channel in our list. Well, there are two reasons for that. First, he addresses a multitude of topics on his channel that are related to the Data Science domain. Some of these topics include Neural Networks, Linear Algebra, Fourier Transformations, Calculus, and many others.
Second, the amazing visualizations you see in his videos have been created by an animation engine in Python called manim, which Grant created by himself. The channel 3Blue1Brown has over 100 uploads with a total of 165 million views.
Throughout his channel, you’ll realize how much of an important role the visualizations play in Grant’s videos and how beautifully the manim library can be utilized to create some slick visuals.

2. freeCodeCamp.org





Source: freeCodeCamp.org

Since: Dec 17, 2014

Creator: Quincy Larson

Views: 127,751,078

Subscribers: 2.72M

Youtube Link: https://www.youtube.com/c/Freecodecamp/featured

Website: https://www.freecodecamp.org/

Created by Quincy Larson in December 2014freeCodeCamp is a non-profit organization with a mission to empower people to code and help others. freeCodeCamp is more of a course-oriented channel run by a highly knowledgeable group of people with a strong background in programming. Their YouTube channel offers informative videos on a wide range of topics, such as Data Structures, JavaScript, Python, Data Science, Machine Learning, Node.js, and has upwards of 127 million total views so far.
freeCodeCamp’s channel has over 1100 videos, and a decent chunk of them are full-on courses with over an hour of content and code sessions at the very least. You can also visit their website to get your hands on more than 6000 tutorials on programming and ethical hacking.

3. Sentdex





Source: Sentdex

Since: Dec 17, 2012

Creator: Harrison Kinsley

Views: 87,960,317

Subscribers: 974K

Youtube Link: https://www.youtube.com/c/sentdex/features

Website: https://pythonprogramming.net/

Created by Harrison Kinsley in December 2012, Sentdex covers several programming topics and technologies such as Machine LearningNatural Language ProcessingData Analysis and Visualization, and some Robotics projects with Raspberry Pi projects.
Harrison’s clear and explanatory style of simplifying the various topics puts Sentdex among the best Data Science channels on YouTube. The channel has over 1200 videos and more than 87 million views. Harrison’s love for Python can be seen on his channel, as he has covered a multitude of programming topics in Python.
He also runs a website called Python Programming Tutorials, where you can find a healthy collection of Python projects in a very detailed manner and see how things work. If you’re interested in a more advanced topic, say, Neural Networks, Harrison has written a book on it called “Neural Networks from Scratch in Python”.
Some Interesting Articles related to Python —

Best Python IDEs and Code Editors You Must Use in 2020

Top Python IDEs and Code Editors with noteworthy features

towardsdatascience.com

10 Cool Python Project Ideas for Python Developers

A list of interesting ideas and projects you can build using Python

towardsdatascience.com

Python Books You Must Read in 2020

Have a look, why you should read them?

towardsdatascience.com

4. Corey Schafer





Source: Corey Schafer

Since: Jun 1, 2006

Creator: Corey Schafer

Views: 47,336,678

Subscribers: 661K

Youtube: https://www.youtube.com/c/Coreyms/featured

Website: http://coreyms.com/

Corey Schafer’s YouTube channel revolves mainly around programming tools that are vital for modern programmers and researchers, including the fundamental concepts of programming. The videos on Corey’s channel have garnered over 47 million views and counting. The channel has covered a range of topics, such as the basics of programming, Linux tutorials, SQL tutorialsDjango, and much more.
For individuals interested in Data Science, Corey has got you covered with video playlists on topics like PandasMatplotlib, and a series of videos on getting started Python. Whether you’re a veteran in programming in a professional environment or a beginner learning about the technology, Corey has offered educational content keeping everyone’s skill level in mind.

5. Tech With Tim





Source: Tech With Tim

Since: Apr 23, 2014

Creator: Tim Ruscica

Views: 32,827,834

Subscribers: 466K

Youtube Link: https://www.youtube.com/c/TechWithTim/featured

Website: https://www.techwithtim.net/

Started by Tim Ruscica in April 2014, the videos on Tim’s channel are more focused on Python programming in general with some game development using PyGamesome tutorials on Machine Learning, and JavaScript paired with a few frameworks. With over 32 million total views, Tim’s channel has some cool projects on a couple of topics, such as a Flappy Bird game, a Face Recognition tool, a Slack bot, and more.
Tim has also done a handful of long coding live streams ranging anywhere from a manageable period of 2 hours to more grueling 12-hour sessions. You will also find some crucial tips and advice on his channel for new developers as well as some programming project ideas, along with some beginner-friendly tutorials on Golang and Flutter. If you like the projects on Tim’s channel, you can find the code on his GitHub repo to follow through with some of his videos.

6. Python Programmer





Source: Giles McMullen

Since: Aug 16, 2008

Creator: Giles McMullen

Views: 7,440,604

Subscribers: 229K

Youtube Link: https://www.youtube.com/c/FlickThrough/featured

Giles McMullen created this channel in August 2008 to inspire the world about Python and to show his love for the programming language, hence the name Python Programmer. Throughout the years, Giles covered loads of tutorials on his channel for various topics ranging from the more fundamental ones for beginners like the basics of Python programming to more advanced topics, like Data Science and Machine Learning.
You can find a few free courses on Giles’ channel for Data Science and Machine Learning, which can essentially give you a strong idea about the core concepts in these subjects. Apart from covering educational topics, Giles also has videos about popular Python libraries, such as Pandas, NumPyScikit-learn, and some handy tips to programmers of all skills on the hows and whats of Python or just programming in general.

7. StatQuest With Josh Starmer





Source: StatQuest

Since: May 24, 2011

Views: 17,551,985

Subscribers: 371K

Youtube Link: https://www.youtube.com/c/joshstarmer/featured

Website: https://statquest.org/

What started as mere explanations about the complex statistical techniques to colleagues at work soon turned into Josh’s passion and later into StatQuest. StatQuest takes away the challenges people usually face in understanding those complicated terms and methods that modern statistics and Machine Learning is filled with.
Created by Josh Starmer in May 2011, StatQuest has more than 180 videos with a combined total of 17 million views. In the channel, you will find a handful of playlists explaining the various fundamental concepts like Logistic Regression, Linear Regression, Linear Models. You can also visit StatQuest’s website to find study guides filled with detailed information for a better understanding of sub-topics like AdaBoost, Classification Trees, and a few others.

8. Krish Naik






Posted by uniqueone
,

https://youtu.be/MMQjmeTmoxQ 

최재식 교수님께서 설명가능한 AI(딥러닝)에 관한 여러 모델을 알기 쉽게 말씀해 주셨습니다.


Posted by uniqueone
,

안녕하세요, TensorFlow KR 여러분!
StarGAN 리뷰 및 코드 실습 영상을 만들어 공유합니다. 한 장의 사진을 업로드해서 성별/나이/머리색 정보를 바꿔볼 수 있습니다. (구글 Colab에서 실행해 볼 수 있도록 만들어 놓았으므로, 관심이 있으신 분은 바로 코드부터 돌려보셔도 됩니다.)

코드 실습: https://github.com/ndb796/Deep-Learning-Paper-Review-and-Practice/blob/master/code_practices/StarGAN_Tutorial.ipynb

StarGAN은 Image-to-image Translation 분야에서 유명하고 오래된 (나온 지 3년 정도 된) 논문으로 TensorFlow KR에 계신 많은 분이 알고 계실 Sung Kim 교수님, 주재걸 교수님이 참여한 논문(최윤제님의)이기도 합니다.

현재는 StarGAN v2가 있으므로... StarGAN v2를 리뷰하려고 했으나 양이 너무 많아질 것 같아서 StarGAN v1부터 리뷰합니다. 워낙 말주변이 없어서... 간단하게 편집도 했는데도 1시간 20분짜리가 나왔네요... 필요하신 분께 도움이 되었으면 좋겠습니다.

https://www.youtube.com/watch?v=-r9M4Cj9o_8

Posted by uniqueone
,