Pro Yearly is on sale from $80 to $50! »

Teaching Machine Learning

Teaching Machine Learning

Insights on how to teach machine learning and deep learning.
The entry barrier is not that high!
Video: https://www.youtube.com/watch?v=dyoxtDhUR74
Conference: https://pydata.org/warsaw2017/

1b324e4900e79878eb518c1263b41795?s=128

Piotr Migdał

October 19, 2017
Tweet

Transcript

  1. Teaching Machine Learning Piotr Migdał, PhD http://p.migdal.pl @pmigdal PyData Warsaw

    Conference 19 Oct 2017
  2. PhD in quantum physics theory (2014, ICFO, Barcelona) data scientist

    (deepsense.ai / consultant)
 machine learning deep learning data-viz (D3.js)
  3. I teach…

  4. None
  5. Outline • Everything is easy • Typical problems • Getting

    pragmatic
  6. None
  7. A person first learns classical mechanics • …by playing with

    balls, blocks? • …by learning Newton laws, differential calculus?
  8. A person first learns natural numbers • …by counting apples,

    toys? • …by the von Neumann construction?
  9. i ~ ˙ = ⇣ ~2 2m r2 + V

    ( x ) ⌘ ˆ H = E
  10. A person first learns quantum mechanics • ...by learning linear

    algebra, complex numbers?
  11. http://p.migdal.pl/2016/08/15/quantum-mechanics-for-high-school-students.html

  12. http://quantumgame.io/

  13. A person first learns machine learning • ...by studying computer

    science, mathematics
 and statistics for years?
  14. you don’t understand Machine Learning unless you can teach it

    with pen&paper (or at least - JavaScript)
  15. None
  16. https://generalabstractnonsense.com/2017/03/A-quick-look-at-Support-Vector-Machines/

  17. Machine Learning https://twitter.com/b0rk/status/821922905890103298/photo/1

  18. Decision trees, visually http://www.r2d3.us/visual-intro-to-machine-learning-part-1/

  19. http://p.migdal.pl/2017/01/06/king-man-woman-queen-why.html Julia Bazińska’s talk “Exploring word2vec vector space" - tomorrow

    at 15:00 Word2vec
  20. Matrix factorizaton http://p.migdal.pl/matrix-decomposition-viz/, work in progress

  21. https://distill.pub/

  22. …but at least Deep Learning is hard, isn’t it?

  23. None
  24. None
  25. None
  26. Spreadsheet-based deep learning http://www.deepexcel.net/

  27. http://setosa.io/ev/image-kernels/

  28. Things which are not problems • I don’t know Python

    • They are only high-school • They are not from STEM
  29. Trypophobia detector by high-school students https://github.com/cytadela8/trypophobia

  30. Trypophobia detector by high-school students https://github.com/cytadela8/trypophobia

  31. Problems with teachers • Too much math details & too

    little insight • Too much historical inertia • No plots • Too little real data
 (e.g. all np.random.randn(n, m))
  32. Real data, plots > arrays

  33. Pragmatic algorithms • kNN • Linear + Logistic
 Regression •

    Random Forest • XGBoost • Neural Networks
  34. Problems with clients • Everyone is an “expert” • Squeezing

    one semester (or a few)
 into a few days (or just one) • Deep learning will solve all our problems • Installation!!!
  35. Logistic Regression
 vs Random Forest vs Deep Learning https://github.com/szilard/benchm-ml http://datascience.la/benchmarking-random-forest-implementations/

  36. Easy setup • Python 3 with Anaconda • Jupyter Notebook

    • Neptune.ML
  37. Thank you! Questions? more on my blog
 http://p.migdal.pl/ @pmigdal