What Problem Are You Trying to Solve, Anyway? - Alfred Lee

What Problem Are You Trying to Solve, Anyway? - Alfred Lee

If you're changing careers into programming, or especially data science, don't be intimidated by the stars and the experts. You may know more than you think you know.

79ecef8e99fbedb7bca755c7ec1926f1?s=128

PyGotham 2014

August 17, 2014
Tweet

Transcript

  1. What Problem Are You Trying to Solve, Anyway? Alfred Lee,

    PhD! Data Scientist, Paperless Post! Instructor, General Assembly! PyGotham 2014
  2. Why am I talking at you? Imposter Syndrome Synthesis Python

    as Playground
  3. Who am I? SOURCE: (NASA photo / Jim Ross) http://www.nasa.gov/

    mission_pages/SOFIA/multimedia/imagegallery/index.html
  4. Who am I?

  5. Who am I?

  6. Who am I?

  7. Who am I?

  8. What’s it like? An illustrative, though atypical example

  9. A problem Let’s cluster our users

  10. The Standard Solutions SOURCE: scikit-learn docs: http://scikit-learn.org/stable/ auto_examples/cluster/plot_cluster_comparison.html

  11. The Question How do our users use the site?

  12. Tangent 1 SOURCE: http://en.wikipedia.org/wiki/Sequence_alignment

  13. Tangent 2 SOURCE: http://en.wikipedia.org/wiki/Disjoint- set_data_structure http://en.wikipedia.org/wiki/Percolation_theory

  14. Tangent 3 SOURCE: https://www.youtube.com/watch?v=HGnpkxrRxZI

  15. Putting it together

  16. User Snippets …

  17. Weighted Jaccard

  18. Nope … …

  19. cosine distance with tf-idf weighting Term frequency - inverse document

    frequency cos θ θ
  20. cluster 1/5

  21. cluster 2/5

  22. cluster 3/5

  23. cluster 4/5

  24. cluster 5/5

  25. What’s it like? Lots of places to synthesize new ideas

    from past experiences
  26. What it seems like the stars know Idols You

  27. Actually, knowledge is fuzzy Idols You

  28. Python A productive playground for prototyping possibilities

  29. Closing Note Don’t wantonly break things, try not to fail.

    Not thinking, not creating, not doing is failing. You don’t have to be perfect to be useful.
  30. Thank you Alfred Lee, PhD Data Scientist, Paperless Post Instructor,

    General Assembly @alphrabet https://www.paperlesspost.com/jobs