Today I Learned

TIL, 2018-05-08. Axios, recompose, thinking about ML.

Musings, JS

  • GET: Don’t send data. You send things via the parameters. Search usually uses GET.
  • How often do you update dependencies? Not really sure lol.
  • Don’t overthink things. Just go with Bootstrap for now.
  • React recompose spinner (withBranch) and for lifecycle hooks.
  • Does this thing have to exist on every damn form? Lol.
handleSubmit = (e) => {

Musings, Ruby

  • Array#bsearch is really faster than find, but the array must be sorted first. (Like for 50M numbers, find takes 3 seconds, and bsearch takes like 0 lol.
  • There was a Ruby Enterprise Edition before? (2012.)
  • Basic Auth
  • Admin panel: Administrate.

Musings, CS

  • Learning Machine Learning:
    • Studying about it and tinkering around with random data sets.
    • A lot of the time, your code would not look nice. The key is, does it work?
    • On math: “Let’s figure this out!”
    • So you basically need linear algebra, statistics, probability, and calculus.
    • Python vs R: Both are good, pick and learn one, and learn the language as a language, not as an ML language. “Data Science from Scratch”, “Machine Learning with R”.
    • ML Learning Pipeline: Find data (scrape, APIs, CSVs).
    • Machine Learning Mastery website.
    • Visualize results. If you can’t communicate your result, you might as well not have done it.
    • scikit and algorithm things.
    • The plan:
      • Learn the math (2-3 months).
      • Learn the programming language (1 month).
      • Machine learning tutorials and test projects (1-2 months).
      • Short-term passion projects (1 month).
    • You’re going to fail a lot.
    • Your models may not be predictive.
    • Track your projects on Github and write up your results.

This project is maintained by daryllxd