Showing only posts by Jon Steven Dal Williams. Show all posts.

Thoughts on Reproducibility in Data Science

Coming from a natural science background, it’s unsurprising that I have some pretty strong opinions on reproducibilty. Most data science textbooks will emphasize the issues of imbalanced data sets, overfitting, and stratification, but this only scratches the surface of potential issues encountered in reproducibility. A great deal of human …

« newer articles | page 2