topic Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1 in Weekly Challenges

Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

AYXAcademy — Mon, 13 Sep 2021 17:49:25 GMT

A solution to last week's challenge can be found here.

We’re celebrating the launch of a new Alteryx podcast, Top Shelf Data Science, in our Weekly Challenges today and next week!

Top Shelf will feature a variety of expert guests who are driving innovation in the field, in conversation with @SusanCS, data science journalist for the Community. You’ll hear technical details of their work and their personal stories of doing “top shelf” data science.

Join in the fun and figure out the details of our “data-driven” cocktail recipe in this challenge, and get the rest of the components and instructions next week. You’ll use a range of tools for these challenges, including some predictive tools. But don’t worry, we’ll be sticking to their default settings. It’s like a wine or whiskey tasting — just a sampling — that we hope both predictive newbies and connoisseurs will enjoy.

Both challenges use a Kaggle dataset of cocktail recipes. You’ll need this PDF of the recipe to fill in as you work. Next week we’ll give you the full answers so you can make sure your cocktail is the perfect blend.

For this week’s challenge, you’ll need to use the market basket (“MB”) tools and the Naive Bayes classifier tool, among others. Here are some resources that might help if these tools are new to you or you’d like a refresher:

Market Basket: MB Rules and MB Inspect tools help documentation; Market Basket 101 and 102 on the Data Science Blog
Naive Bayes Classifier: help documentation

Remember, we’re not trying to tune and refine models here; we’re just getting a taste for how the predictive tools work. So you can leave the tools’ settings on their default options, unless otherwise directed, and use the results you obtain to complete the recipe.

Finally, if you haven’t yet done so, you’ll need to install the R predictive tools for this challenge and next week’s challenge.

Prep that cocktail shaker, fire up your favorite podcast app, and get ready for an educational, enlightening happy hour!

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Qiu — Mon, 08 Feb 2021 22:30:59 GMT

This is something new, which I am always keeping😂 distance with..

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Spalders — Tue, 09 Feb 2021 12:27:38 GMT

I think I am ready for Part 2... (however this is certainly new territory)

1. {Powder sugar}->{Lemon}
2. Confusion matrix for the lemon model

3. Mesure questions

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

OllieClarke — Tue, 09 Feb 2021 16:34:09 GMT

Fun challenge - can't wait til next week 🙂

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

phottovy — Tue, 09 Feb 2021 17:24:45 GMT

Excited to see next week!

Q1:

Q2: (I'm not sure I used the oversample tool as intended)

Q3:

PDF (so far?)

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

LHolmes — Wed, 10 Feb 2021 00:53:40 GMT

Super fun to utilize some of the predictive and market basket tools. Can't wait to finish off the recipe for next week.

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

kelvin_law1 — Wed, 10 Feb 2021 01:51:12 GMT

Here is my workflow and answers:

Question 1:

Question 2:

Question 3:

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Kenda — Wed, 10 Feb 2021 02:28:34 GMT

Thankfully, after reading the instructions thoroughly, it didn't end up being as intimidating as it first seemed.

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

atcodedog05 — Thu, 11 Feb 2021 08:27:01 GMT

Market Basket analysis is definitely new segment. This was a great exposure.

Cheers and Happy Analyzing 🙂

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

BeginnerMindset — Thu, 11 Feb 2021 16:41:57 GMT

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Renato_7 — Fri, 12 Feb 2021 15:15:57 GMT

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

TonyA — Fri, 12 Feb 2021 21:28:05 GMT

My first time using MB and Bayesian Classification. This was a great introduction.

Another first for me. The first time I ever voluntarily switched to a vertical layout This workflow seemed to beg for it. The horizontal layout was too tall and narrow.

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Tyro_abc — Sat, 13 Feb 2021 06:19:06 GMT

1
2
3a	strMeasure1 2 oz
3b	strMeasure1 1/2 oz
3c	strMeasure1 1 tsp superfine
3d	strMeasure1 3 oz

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

Qiu — Sat, 13 Feb 2021 08:09:59 GMT

It is learning process

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

AkimasaKajitani — Sat, 13 Feb 2021 10:34:58 GMT

My solution.

I am not confident that my prediction using oversampling is correct.

Result :

Result 1 :

Result 2:

Add Lemon Juice.

Result 3:

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

mbogusz — Sat, 13 Feb 2021 16:19:55 GMT

Great intro to the MB Tools and NB Classifier Tool too

Challenge #254 in the can

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

JP_SDAK — Mon, 15 Feb 2021 18:27:13 GMT

I've never used these tools, this is a great challenge to try all of them.

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

patrick_digan — Tue, 16 Feb 2021 13:41:48 GMT

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

RolandSchubert — Tue, 16 Feb 2021 17:09:00 GMT

Re: Challenge #254: DATA-DRIVEN COCKTAIL RECIPE - PART 1

JennyMartin — Wed, 17 Feb 2021 08:43:29 GMT

Had to peek at the solution to check I was along the right lines here! Lot's of tools I rarely use in this challenge. Video solution here: https://www.youtube.com/watch?v=lSr-CYVxEdw&list=PL18SpOXDj-KMUD4YwWC5vZaUsIa7drxNZ&index=58