Analytics Blog

News, events, thought leadership and more.
Register for the upcoming Live Community Q&A Session - and don't forget to submit your questions for @DeanS regarding the future role of analytics here.
Alteryx
Alteryx

What’s New?

The Assisted Modeling teams have been working hard to bring some awesome new features to the Assisted Modeling experience in Designer! We hope that these new features will provide our customers further insights into the Assisted Modeling journey and offer more opportunities for our customers to customize their predictive analytics workflows to meet their specific needs.  We have added the model train time to the Assisted Modeling leaderboard to help our customers make decisions about which model best suits their needs. We have also added a button next to each model in the Assisted Modeling leaderboard that allows you to output a model report and output Python code into the Designer Python tool!

BradW_1-1588094325396.png

figure 1 - Export code into Designer Python tool!

 

Why’s it Cool?

With our new metrics for examining training time in the leaderboards, you can now determine which model to use based on the expense of training the model, so you don’t have to sacrifice your computation time for quality in your Assisted Modeling Pipelines.

 

With the new HTML Report Output you can now save the results of your machine learning model and capture the metrics in a clean, elegant markup file. Use this feature to track changes in your models while tweaking parameters to help find the best predictive model for your use-case! After you’re done, export the report output to share your model results with others in the organization.

 

Do you want to understand the “how” behind your predictive power in Designer? Assisted Modeling has enabled you to dive into machine learning as deep as you would like! With the new Python Code Output functionality in Assisted Modeling, you can export your predictive pipeline to the Python Tool. From there you can easily understand, customize and optimize your Modeling Pipeline to meet your every particular need. The seamless transition from Assisted Modeling to Python Tool allows you to synchronize the steps you take in Assisted Modeling with the Python code you are using behind the scenes.

BradW_2-1588094325402.png

figure 2 - Modeling Pipeline in Alteryx Designer

 

 

What’s Next?

Try it out and give us your feedback.  Existing beta participants can sign in here (content highlighted in screenshot below).  Not a beta participant, sign up here (beta participation requirements here).  Questions?  betas@alteryx.com

 

Picture1.png

figure 3 - Assisted Modeling content in Alteryx Beta Program

 

Comments
11 - Bolide

Thanks @BradW 

 

Can you help find resources on Modeling in Alteryx on the community?

 

Hope you are safe,

 

J

Alteryx Community Team
Alteryx Community Team

@the_jake_tool 

  1. First, check out the Data Science learning path
  2. For more on Assisted Modeling, check out

     Is Assisted Modeling for Me? and

     What's a Pipeline? An Overview of the New Python-based Machine Learning Tools

  3. For more on the predictive tools, check out the Data Investigation, Predictive, Time Series, Predictive Grouping, and Prescriptive sections of the Tool Mastery series
  4. Finally, subscribe to the Data Science blog!