Back to archive

Thread

2 tweets

1
To people complaining that data science is 80% data cleaning, it's called DATA science not model science ;) (And yeah, it's painful at times).
2
Also, data cleaning is the wrong image, I think. Often it's not just about cleaning but figuring out if it is the right kind of data, has the information you need, etc. and then actively looking for better data. It's not like you get a dirty slump of data and that's it.