Challenge #255: DATA-DRIVEN COCKTAIL RECIPE - Part 2

[TonyA]

Had to bang my head on a couple of these for a while, but got it done. Probably learned more from this challenge than any other in the past six months.

Spoiler

#6 was a little tricky. I'm always telling new users how important it is to profile data before jumping in. I forgot my own advice this time and it cost me.

[kelvin_law1]

I have combined the workflow for both Challenge 244 & 245:

Spoiler

[Tyro_abc]

[Spalders]

Spoiler

Spoiler

Cheers!

[PhilipMannering]

Here's my solution. I was a bit confused why everyone else was getting soda and I was lacking. Needed to do do a chi-square test, not a t-test.

Spoiler

workflow

[Sashikumar]

Solved.

Spoiler

[Kenda]

Spoiler

[BeginnerMindset]

Spoiler

Spoiler

 Update: 19/02/21

I couldn't resist having more of a play with the Random Forest model.  If I use the ingredients I can reduce the classification error for "stir" as well as the overall OOB error.

Spoiler

Spoiler

[mbogusz]

Shout out to Mothers Against Drunk Driving: MADD

Spoiler

Challenge #255 in the can... how was this different than #254???

[glenperes]

Solution 254 and 255