Skip to content

bydmitry/HSE_deeplearning

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

HSE_deeplearning

Fork of Lempitsky DL for HSE master students

Lecture and seminar materials for each week is in ./week* folders

Binder (servers may be down time to time, sry)

Coordinates

  • Autumn'16 track finished, new track eta next autumn. Meanwhile, course materials will be reused for YSDA.
  • Any technical issues, ideas, bugs in course materials, contribution ideas - add an issue

Announcements

  • 27.12 - autumn'16 track finished
  • 23.12 - added some bonus links in most weeks for those mysterious people who lurk this repo from outside the HSE.
  • 23.12 - pre-exam Q&A happened. Exam will happen next monday at 16.40, room 311
  • 17.12 - added MIT just for kicks
  • 16.12 - exam/project rules published - https://github.com/yandexdataschool/HSE_deeplearning/wiki/How-exam-works - ping us on which problems you prefer (see slack for reservations; will add table tomorrow)
  • 16.12 - cheched up whatever you sent us before 16.00 PM 16.12. Some of your submissions haven't been graded yet? ping us asap.
  • 16.12 - bayesian lecture happened
  • 13.12 - A note to HSE students - final deadline is department-forced to happen on 26.12. We are [probably] terribly sorry. Please make sure that your points so far is at least 40 or will get there in time. We'll also have a practical exam, with tasks distributed next lecture. Details
  • 13.12 - another total chechup wave (everything you sent us by 01:30 13.12.2016 should have been checked. If not, ping us/create issue)
  • 9.12 - week11 uploaded
  • 2.12 - updated info
  • 2.12 - checked up all submissions sent before 5 AM (we hope). If yours is missing, PM or issue us.
  • 2.12 - winter is coming... winder is here. Deadlines are coming!
  • 25.11 - week9 by Arseny Ashukha on deep learning for sound processing
  • 18.11 - week8 by Dmitry Ulyanov
  • 15.11 - HW checkup wave. If you sent us your homework solution before 05:00 15.11.2016 but we never replied yet - go get us @slack or post an issue here - we'll help
  • 13.11 - we fix__d__ week7 assignment and uploaded all the PDFs. Homework checkup wave is underway :)
  • x.11 - another wave of homework checkups happened
  • 3.11 - week7 will happen on 11.11.2016 the regular way (friday, 18-00, room 400)
  • 3.11 - we checked up all homework assigments sent to us before 4:00 am 3.11.16. If you were for some reason left behind - write us (mail, slack, github issue, irl, ...) - we'll fix that.
  • 3.11 - week6 homework announced
  • 1.11 - Week6 lecture will occur on Wednesday, 2.11 at 18-00 at approximately the same room 400.
  • 26.10 - Preliminary poll on next lecture date: http://doodle.com/poll/cvew2rkn9u9x5ept . Please participate ASAP.
  • 24.10 - Checked up homeworks "so far". If you sent us any mail@homework before 4.20 24.10.16 and we haven't replied yet - PM us in slack or create github issue or just contact us IRL - we'll fix that.
  • 21.10 - There will be no classes on the 28.10;
  • 21.10 - week5 notebook is NOT homework assignment notebook. Homework and deadlines TBA
  • 20.10 - we now have a Binder for you if you got to the seminar without a notebook (with a phone) or have some temporary technical issues. Binder only lasts 1 hour so please do not use it for homeworks.
  • 20.10 - current status of course staff is "Адъ и израиль", so the new wave of homeworks will be checked with a short delay (hope to finish by the weekend. The lectures will proceed as planned.
  • 17.10 - week4 partially completed seminar uploaded
  • 17.10 - a way for students to reduce lateness penalty on their homework assignments announced
  • 14.10 - week4 lecture, seminar and homework uploaded
  • 12.10 - If you sent us anything before 6-00 AM (Moscow) 12.10.16 and still got no reply / no score here - contact us (slack, issue on github, anything) - there may be a problem with e-mail delivery.
  • 12.10 - for those rare specimen who read official curriculum - we'll have have to reorder the curriculum. Advanced vision goes 2 weeks forward, advanced text gets 2 weeks sooner. The rest stays as planned.
  • 7.10 - If you are considering project ideas, please contact us (slack >= mail) asap to know you exist. You don't have to know actual project topic - just tell us that you're there and you may be up to something.
  • 7.10 - We've now got da feedback form - it's fully anonymous and you can send there whatever you won't send us via slack/mail.
  • 7.10 - Week3 lecture notes and homework uploaded.
  • 30.09 - Week2 homework (lasagne/cifar): Please do not forget to add deterministic=True for your neural network when computing accuracy (not when training)
  • 30.09 - added week2 lecture and homework.
  • 30.09 - published scoreboard
  • 23.09 - added week1 materials and homework
  • 20.09 - by default we meet on Friday at 18-00, room 400. if you cannot make it, please send us an e-mail (course mail) or PM me in slack ASAP - we have a second parallel available and a few other options.
  • 20.09 - HW0 deadline was shifted 1 wek into the future. Rejoice!
  • 20.09 - grading info added
  • 15.09 - Doodle on when do we meet - link
  • 15.09 - Please get the frameworks installed by the next class - issue
  • 15.09 - added project rules and examples (see "course stuff" below)
  • 15.09 - week0 assignment published (see "syllabus")

Syllabus

  • week0 Recap
    • Lecture: Linear models, stochastic optimization, regularization
    • Seminar: Linear classification, sgd, modifications
      • HW due: 28.09.16, 23.59.
    • Please get bleeding edge theano+lasagne installed for the next seminar.
  • week1 Getting deeper
    • Lecture: Neural networks 101
    • Seminar: theano, symbolic graphs and basic neural networks
      • HW due: 3.10.16 23.59
  • week2 Deep learning for computer vision 101
    • Lecture: Convolutional neural networks
    • Seminar: lasagne and CIFAR
      • HW due: 9.10.16 23.59 on first submission.
  • week3 Deep learning for natural language processing 101
    • Lecture: NLP problems and applications, bag of words, word embeddings, word2vec, text convolution.
    • Seminar: Text convolutions for Avito content filtering task
      • HW due: 16.10.16 23.59 on first submission.
  • week4 Recurrent neural networks for sequences
    • Lecture: Simple RNN. Why BPTT isn't worth 4 letters. GRU/LSTM. Language modelling. Optimized softmax. Time series applications.
    • Seminar: Generating laws for pitiful humans with mighty RNNs.
      • HW due: 28.10.16 23.59 on first submission.
  • week5 Recurrent neural networks II
    • Lecture: Batchnorm and dropout for RNN; Seq2seq: machine translation, conversation models, speech recognition and more. Attention. Long term memory architectures.
    • Seminar: a toy machine translation task
      • to be anounced
  • [Skip week]
  • week6 Fine-tuning with neural networks
    • Lecture: Large CV datasets, model zoo, reusing pre-trained networks, fine-tuning, "knowledge transfer", soft-targets
    • Seminar: Cats Vs Dogs Vs Very Deep Networks
      • HW due 17.11.16 23.59
  • week7 Advanced computer vision
    • Lecture: Representations within convnets, fully-convolutional networks, bounding box regression, maxout, etc.
    • Seminar: Image captioning by Arseniy Ashukha
      • HW due 24.11.16 23.59
  • week8: Generative models for computer vision
    • Lecture: Autoencoders, Generative Adversarial Networks
    • Seminar: Art Style Transfer with deep learning (Dmitry Ulyanov)
      • HW due 4.12.16 23.59
  • week8: Deep learning for sound processing
    • Lecture: case study: music recommendation with deep learning
    • Seminar: Music clustering & content-based recommentation with convolutional nets
      • HW due 11.12.16 23.59
  • week9: Basic reinforcement learning
    • Lecture: Introduction to reinforcement learning
    • Seminar: one algorithm to navigate in a maze, play pacman and control robots.
      • HW due 11.12.16 23.59
  • week10: Deep reinforcement learning
    • Lecture: approximate reinforcement learning with deep neural networks (problems and solutions)
    • Seminar: Playing Atari/Doom with deep reinforcement learning
      • HW due 18.12.16 23.59 first submission
  • week12: Bayesian deep learning
    • Lecture: Basics of bayesian approach to probabilities
    • Bonus lecture: Variational autoencoders (Mikhail Khalman)
      • HW due 26.12.16 16.00 hard

Stuff

Contributors & course staff

Course materials and teaching performed by

About

Fork of Lempitsky DL for HSE master students.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 87.2%
  • Python 12.8%