A friend of mine learned data science on the job with real, messy datasets. When they finally took an intro to ML course and learned how easy it was to work with Irises, they were like, "Well, I probably would have been really mad if it'd happened in the opposite order." https://t.co/BiI1GWV3gT
-
This Tweet is unavailable.
-
Replying to @imightbemary
I learned on the job with literal production logs... so whenever someone hands me a clean dataset I essentially start asking "what are you hiding from me?"pic.twitter.com/BkrN8SbvzM
1 reply 0 retweets 6 likes -
-
Replying to @imightbemary @Randy_Au
I secretly kind of enjoy messy data, as long as it’s the kind of messy that can be cleaned up.
1 reply 0 retweets 1 like -
Replying to @SamanthaZeitlin @Randy_Au
I feel this. Data cleaning is a DS task that there's a clear endpoint to. The feeling of unambiguously finishing a task is a harder endorphin rush to come by in DS than in a lot of other tech professions
1 reply 0 retweets 0 likes -
Replying to @imightbemary @Randy_Au
Having spent the majority of my career in research, I don’t know that I’ve ever unambiguously finished a task in my life. You’re never really done. I just happen to subscribe to the school of thought that data cleaning is a form of analysis.
1 reply 0 retweets 0 likes
Agreed on both accounts. You're never done with analysis work because the systems you're studying (and your understanding of them) are constantly evolving. As for cleaning, I see it as being about honing in on the correct abstractions.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

