태그 보관물: NIPS2016

Reliable ML in the Wild – NIPS 2016 Workshop

댓글 남기기

NIPS 2016 워크샵에서 관심을 많이 받았던 것 중에 하나인 ‘Reliable Machine Learning in the Wild‘의 동영상이 워크샵 스케줄 페이지에 링크되었습니다. 비디오가 공개된 것만 추려보았습니다. 비디오는 wmv 파일로 스트링이 아니고 다운로드입니다. 전체 동영상을 한꺼번에 받으시려면 여기에서 전체 목록을 쉽게 확인할 수 있습니다. 동영상외에 전체 자료는 워크샵 스케줄 페이지를 확인하세요.

8:40 – Opening Remarks (video)

Session 1: New Challenges in Machine Learning

9:00 – Invited talk: Rules for Reliable Machine Learning. Martin Zinkevich (video)

9:30 – Contributed talk: What’s your ML Test Score? A rubric for ML production systems. Eric Breck, Shanqing Cai, Eric Nielsen, Michael Salib, D. Sculley (slides) (video)

9:45 – Poster spotlights (video)

Session 2: Robustness to Model Mis-specification

10:30 – Invited talk: Robust Learning and Inference. Yishay Mansour (video)

11:00 – Invited talk: Automated versus do-it-yourself methods for causal inference: Lessons learned from a data analysis competition. Jennifer Hill (video)

11:30 – Contributed talk: Robust Covariate Shift Classification Using Multiple Feature Views. Anqi Liu, Hong Wang Brian D. Ziebart (video)

11:45 – Poster spotlights (video)

Session 3: Machine Learning and Security

1:30 – Contributed talk: Learning from Untrusted Data. Moses Charikar, Jacob Steinhardt, Gregory Valiant (slides) (video)

1:45 – Invited talk: Adversarial Examples and Adversarial Training. Ian Goodfellow (video)

2:15 – Contributed talk: Summoning Demons: The Pursuit of Exploitable Bugs in Machine Learning. Rock Stevens, Octavian Suciu, Andrew Ruef, Sanghyun Hong, Michael Hicks, Tudor Dumitras (slides) (video)

2:30 – Poster spotlights (video)

Session 4: Reliable Robotics in Complex Environments

3:30 – Invited talk: Learning Reliable Objectives. Anca Dragan (video)

Session 5: Reliable Machine Learning in Practice

4:30 – Contributed talk: Online Prediction with Selfish Experts. Okke Schrijvers (video)

4:45 – Contributed talk: TensorFlow Debugger: Debugging Dataflow Graphs for Machine Learning. Shanqing Cai, Eric Breck, Eric Nielsen, Michael Salib, D. Sculley (slides) (video)

Rules of ML: Best Practice for ML from Google

댓글 남기기

구글 연구원인 Martin Zinkevich 가 만든 ‘Rules of ML: Best Practices for ML Engineering‘ 이 공개되었습니다. 이 파일은 지난 NIPS 2016의 Reliable Machine Learning in the Wild 워크숍에서 발표한 내용을 정리한 것으로 보입니다. 늘상 보아오던 가이드라인이 아니라 실제 구글에서 여러 서비스를 만들면서 얻은 노하우를 사례를 들어가며 설명하고 있습니다. 이번 주말에 보아야할 문서로 안성맞춤입니다! 🙂

NIPS 2016 GAN Tutorial Summary

댓글 남기기

nips_2016_gan_report

OpenAI 의 이안 굿펠로우(Ian Goodfellow)가 NIPS 2016의 GAN 튜토리얼을 요약한 리포트를 만들어서 Arxiv 에 등록하였습니다. 생성 모델(generative model)이 인기를 끄는 이유와 어떻게 작동하는지, GAN(Generative Adversarial Network)이 다른 모델과 다른 점과 GAN이 작동하는 상세 내용을 다룹니다. 또 최근 GAN 연구 동향과 최신 모델도 함께 다루고 있어 놓치기 아까운 리포트인 것 같습니다!

Repo. for NIPS 2016 papers

댓글 남기기

NIPS 2016 페이퍼가 구현된 레파지토리를 정리하는 레딧 포스트가 있습니다. 지금까지 13개의 깃허브 레파지토리가 정리되어 있습니다. 추가되는 대로 업데이트 하겠습니다. 혹시 이 외에 다른 레파지토리가 있다면 공유 부탁 드립니다.

  1. Using Fast Weights to Attend to the Recent Past (https://arxiv.org/abs/1610.06258)
    Repo: https://github.com/ajarai/fast-weights
  2. Learning to learn by gradient descent by gradient descent (https://arxiv.org/abs/1606.04474)
    Repo: https://github.com/deepmind/learning-to-learn
  3. R-FCN: Object Detection via Region-based Fully Convolutional Networks (https://arxiv.org/abs/1605.06409)
    Repo: https://github.com/Orpine/py-R-FCN
  4. Fast and Provably Good Seedings for k-Means (https://las.inf.ethz.ch/files/bachem16fast.pdf)
    Repo: https://github.com/obachem/kmc2
  5. How to Train a GAN
    Repo: https://github.com/soumith/ganhacks
  6. Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences (https://arxiv.org/abs/1610.09513)
    Repo: https://github.com/dannyneil/public_plstm
  7. Generative Adversarial Imitation Learning (https://arxiv.org/abs/1606.03476)
    Repo: https://github.com/openai/imitation
  8. Adversarial Multiclass Classification: A Risk Minimization Perspective (https://www.cs.uic.edu/~rfathony/pdf/fathony2016adversarial.pdf)
    Repo: https://github.com/rizalzaf/adversarial-multiclass
  9. Unsupervised Learning for Physical Interaction through Video Prediction (https://arxiv.org/abs/1605.07157)
    Repo: https://github.com/tensorflow/models/tree/master/video_prediction
  10. Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks (https://arxiv.org/abs/1602.07868)
    Repo: https://github.com/openai/weightnorm
  11. Full-Capacity Unitary Recurrent Neural Networks (https://arxiv.org/abs/1611.00035)
    Repo: https://github.com/stwisdom/urnn
  12. Sequential Neural Models with Stochastic Layers (https://arxiv.org/pdf/1605.07571.pdf)
    Repo: https://github.com/marcofraccaro/srnn
  13. Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering (https://arxiv.org/abs/1606.09375)
    Repo: https://github.com/mdeff/cnn_graph
  14. Interpretable Distribution Features with Maximum Testing Power (https://papers.nips.cc/paper/6148-interpretable-distribution-features-with-maximum-testing-power.pdf)
    Repo: https://github.com/wittawatj/interpretable-test/
  15. Composing graphical models with neural networks for structured representations and fast inference (https://arxiv.org/abs/1603.06277)
    Repo: https://github.com/mattjj/svae
  16. Supervised Learning with Tensor Networks (https://arxiv.org/abs/1605.05775)
    Repo: https://github.com/emstoudenmire/TNML
  17. Fast ε-free Inference of Simulation Models with Bayesian Conditional Density Estimation: (https://arxiv.org/abs/1605.06376)
    Repo: https://github.com/gpapamak/epsilon_free_inference
  18. Bayesian Optimization for Probabilistic Programs (http://www.robots.ox.ac.uk/~twgr/assets/pdf/rainforth2016BOPP.pdf)
    Repo: https://github.com/probprog/bopp
  19. PVANet: Lightweight Deep Neural Networks for Real-time Object Detection (https://arxiv.org/abs/1611.08588)
    Repo: https://github.com/sanghoon/pva-faster-rcnn
  20. Data Programming: Creating Large Training Sets Quickly (https://arxiv.org/abs/1605.07723)
    Repo: https://github.com/HazyResearch/snorkel
  21. Convolutional Neural Fabrics for Architecture Learning (https://arxiv.org/pdf/1606.02492.pdf)
    Repo: https://github.com/shreyassaxena/convolutional-neural-fabrics
  22. Value Iteration Networks in TensorFlow (https://arxiv.org/abs/1602.02867)
    Repo: https://github.com/TheAbhiKumar/tensorflow-value-iteration-networks

NIPS 2016

댓글 남기기

(업데이트) NIPS 컨퍼런스와 튜토리얼의 동영상이 스케줄 페이지에 등록되었습니다. 곧 워크샵 동영상도 제공될 예정이라고 합니다. (2017-02-22)‘Adversarial Training’ 워크샵의 고화질 영상이 유투브에 올라왔습니다.

NIPS 2016 으로 스페인 바르셀로나가 시끌벅적합니다. 많은 내용들이 쏟아져서 다 살펴보기가 어렵네요. 정리가 되는 대로 이 포스트에 자료를 업데이트 하도록 하겠습니다.

NIPS 2016 에 대한 스케줄은 이곳에서 모두 확인할 수 있습니다.

먼저 오늘 열릴 워크샵 중 Adversarial Training이 페이스북 라이브로 중계될 예정입니다. 우리 시간으로 오후 5시 부터 생중계 될 예정입니다. 만약 바르셀로나 현지 네트워크가 원할하지 못해 라이브 스트림이 힘들다면 녹화해서 며칠내로 올린다고 합니다. 이안 굿펠로우(Ian Goodfellow)부터 얀 리쿤(Yann LeCun) 박사까지 발표자가 쟁쟁해서 놓칠 수 없는 이벤트인 것 같습니다.

  • Ian Goodfellow(OpenAI), Generative Adversarial Networks(GANs), slide, video
  • Soumith Chintala(Facebook), How to train a GAN?, github
  • Arthur Gretton(University College London), Learning features to compare distributions, slide, video
  • Sebastian Nowozin(Microsoft), f-GAN, slide, paper
  • Aaron Courville(University of Montreal), video
  • Yann LeCun(Facebook), Energy-Based GANs, slide
  • Panel Discussion, video

요수아 벤지오(Yoshua Bengio) NIPS 2016 슬라이드

개별 슬라이드

  • John Schulman(OpenAI), Deep RL Workshop, The Nuts and Bolts of Deep RL Research, slide
  • Andrew Ng(Baidu), Nuts and Bolts of Building Applications using DL, slide
  • Yann LeCun, Invited Talk, Predictive Learning, slide
  • David Blei/Shakir Mohamed/Rajesh Ranganath, Variational Inference, slide
  • Pieter Abbeel/John Schulman, Deep RL Through Policy Optimization, slide

(업데이트)