How-to-fail-as-Data-Scientist-by-Kyle.txt

Kyle McKiou
Kyle McKiou
I Help Data Scientists Get Jobs! | Director of Data Science | Entrepeneur
4h • Edited
How to Fail Data Science Case Studies

Here are a few mistakes that YOU should avoid if you want to pass take-home assignments:

	1. Write a script that runs from start to finish with no control statements, no functions, and no classes
	2. Don’t tune your models
	3. Blame the algorithms for poor results
	4. Give all your variables cryptic names: x11, x12, x13
	5. Don’t visualize the data
	6. Skip EDA
	7. Don’t even think about feature engineering
	8. Use R... when the directions clearly say “use Python”
	9. Ignore the directions completely
	10. Take 3 weeks to finish because “you’re too busy” (even though you’re unemployed and looking for a job)
	11. Send multiple emails asking questions that are answered in the directions
	12. Don’t include your results in the write up
	13. Don’t spell-check your report
	14. Be incredibly inefficient (e.g. grid search on a massive parameter space) and then blame your computer for “taking too long”
	15. Don’t look for duplicates in the data
	16. Only use advanced models without trying anything simple
	17. Give an incorrect summary of the algorithm you selected
	18. Blatantly copy-paste code off the internet 

Avoid these and you'll have a lot more success with your case studies!

#datascience

And yes, I've seen each of these multiple times!!!