Science Research Weekly

Episode 39: Last Stat Before Christmas

Listen Later

In this episode, I pondered the problem of false positives in publications, rotated around the JuliaHub, machinated over machine learning projects, and put R package ‘drawCell’ to the test.

References:

  • Are most published research findings false in a continuous universe?
  • Principal component analysis
  • A no-free-lunch theorem for multitask learning
  • JuliaHub
  • Signal-to-noise ratio and statistical tests
  • Large Enough Sample Condition
  • Top 11 Interesting Machine Learning Project Ideas For 2023
  • Code longevity of the R programming language
  • {drawCell} – R package and Shiny app for drawing cell structures

    • R-packages:

    • colorRamp2: Generate Color Mapping Functions
    • rworkflows: Test, Document, Containerise, and Deploy R Packages
    • MetChem: Chemical Structural Similarity Analysis
    • panelsummary: Create Publication-Ready Regression Tables with Panels
    • arf: Adversarial Random Forests
      • ...more
      View all episodesView all episodes
      Download on the App Store
      Science Research WeeklyBy Mark R Williamson