/rect-learn-clt

Interactive theoretical and experimental exploration of learning axis aligned rectangles under the PAC learning model.

Primary LanguageJupyter Notebook

rect-learn-clt

Interactive theoretical and experimental exploration of learning axis aligned rectangles under the PAC learning model.

TODO

  • Sliders for adjusting sample size
  • Choose multiple distributions
  • Continue with classic proof for "tightest-rectangle" hypothesis
    • Include visualizations for weight of error under distribution
    • Include sliders for \epsilon and radio buttons for Distribution to see how it effects visuals and experimental computations.