cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Lessons from a Kaggle Winner
Created: Mar 31, 2020 05:17 PM  |  Last Modified: Jun 5, 2020 11:10 AM
kaggle_covid19.jmp

(view in My Videos)

 

Kaggle is the world's largest data science community, bringing together tools, data and thousands of talented data scientists to help solve some of the world’s most complicated and pressing problems. In this segment, Anne Milley interviews Russ Wolfinger, Director of Scientific Discovery at JMP and a Kaggle Grandmaster to learn his secrets of winning Kaggle competitions. Wolfinger is currently participating in a new rapid-response weekly competition series running in community. In Lessons from a Kaggle Champion, Russ will discuss the competition and demonstrate how he’s using the new XGBoost add-in for JMP Pro to accurately forecast coronavirus trends worldwide using only three basic features. Russ’ presentation is the featured segment of the April 6th episode of JMP On-Air.

 

If you would like to try out what you learned from the data science superhero pop over and browse some recent Kaggle data.

Comments
atlas100
‎04-06-2020 11:53 AM

Russ mentions using the XGBoost add-in in this segment. I can't find the add-in on the site. Where is it? Is it a JMP Pro only add-in? 

Peter_Hersh
‎04-06-2020 05:27 PM

Yes, the add-in is only available with JMP Pro as it runs off the Gradient Boosted Tree that is a JMP Pro feature.  If you are interested in trying out the add-in reach out to @Jason_Wiggins he will be able to add you to the list.

PatrickGiuliano
‎04-19-2020 02:48 AM

@russ_wolfingerYou mentioned you would share the Covid cumulative death rate/ infection rate data table you were working with where you demonstrated some operations on the columns (log and lag by 7 says) and then demonstrated the graphing and eventually the use of XGBoost for prediction? 

 

I'd love to play with the dataset and use JMP's 'default' predictive modeling platforms as a learning exercise for myself (in addition to using the XGBoost add-in, which to which I've reached out to @Jason_Wiggins for the Pro add-in as I understand it is only available for Pro at this time.  

russ_wolfinger
‎04-19-2020 09:01 AM

Hi @PatrickGiuliano , thanks, the table is an attachment just above the video.  It's a couple of weeks old though now.  For the latest, check out  https://www.kaggle.com/c/covid19-global-forecasting-week-4/overview  

 

I just invited you to the XGBoost Early Adopter group.  Once in there, click the Welcome message and follow the provided instructions.   

 

@all This group is still open to anyone interested, please message me or @mia_stephens if you would like to join.

PatrickGiuliano
‎04-19-2020 03:37 PM

Thanks @russ_wolfinger! 

 

I see it now in the Kaggle data hyperlink that you included above and directly below the posted video. I re-copy the latest data query link here for my own reference and in service of the community: https://www.kaggle.com/c/covid19-global-forecasting-week-4/overview