Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!

General Discussions

Discuss any topics that are not product-specific here.

Where do you begin with predictive analytics?

LeahK
Alteryx Alumni (Retired)

GOT THOUGHTS.pngThis week’s question came to me while attending a local Alteryx User Group meeting in Denver yesterday evening. At one point during @KatieH's road-map presentation, she casually polled the group and asked how many people were actively using predictive tools...  I think two, maybe three, people raised their hands.

 

No one seemed all that surprised by the lack of response or enthusiasm, and as Katie later shared with the group, it is one of the reasons Product Management is working hard to improve the barrier to entry and overall usability of predictive tools (yay!).


In the meantime, I figured this was a great opportunity to leverage this week’s Thursday Thought to capture the advice and recommendations of those who actively use what’s currently available in Alteryx for predictive analytics.

 

As a reminder, the Thursday Thought is a weekly Q&A Thread where we invite members of our Community to share their experience. 

 

Question:

How would you recommend someone who doesn't have experience using Alteryx for predictive analysis get started? Where should they begin?

2 REPLIES 2
LeahK
Alteryx Alumni (Retired)

I should point out that our community is chock-full of some great resources that can help newer users get started with predictive tools :)

 

MarqueeCrew
20 - Arcturus
20 - Arcturus

@LeahK,

 

Where to start with predictive analytics?

 

I'd recommend rolling up your sleeves and get into the data.  Discovery will help you to understand the content and possibly the context of the data.  There is likely going to be dirty data that will need a little TLC and missing data that you'll need to address.  Whether you're using Alteryx or another tool, this is where everyone will have to start.  If you're used to using another tool and you have access to it, you might want to work in parallel and become comfortable with Alteryx.  If you're using R as your starting point, you'll be in luck and you can easily transition or copy your work into Alteryx.

 

There are many training opportunities to search for concerning predictive analytics, including inspire2017.  I just took a course on grouping analysis the other day and now realize all that I don't know about cluster analysis.  If I get an assignment in that area, I'll at least know where to start and have some leads on finding information both inside and outside of Alteryx.  You might also want to invest in yourself and try Udacity.

 

This morning I posted to a question on the community regarding using Alteryx to read emails and determine if the content of the email indicated acceptance or rejection.  This is a form of predictive analytics.  The user has a list of APPROVAL words and REJECTION words.  Once the email is read, it is desired that the email can be categorized.  My suggestion is that once formally categorized (the dependent variable is set as 0 or 1, Yes or No, Approve or Reject) that the process could construct a predictive model for acceptance or rejection.  I really appreciate and like the offer, BUT i'm NOT interested at this time.  It's me, not you.

 

If you have a problem to solve and you need a little help, check with the community.

 

Cheers,
Mark

Alteryx ACE & Top Community Contributor

Chaos reigns within. Repent, reflect and restart. Order shall return.
Please Subscribe to my youTube channel.
Labels