How do you know which modeling methods will work best on your data? To have the best chance at success you need to build multiple predictive models, compare them and select the best of the best – or even a combination. In our February 25, Technically Speaking event, Kemal Oflus demonstrated how to use machine learning methods without having to write and tune algorithms using your subject matter expertise – and ultimately making better decisions, faster.
The event attracted more than 1,300 registered attendees and secured more than 400 new contacts in Salesforce. As our JMP Sales teams prepare to engage in post-event discussions with registered attendees, we suggest you consider positioning some (or all) of the following resource recommendations.
If you have any questions about these recommendations, please contact me at x11276.
Thanks
Larry
February 25, 2021, Build and Choose Better Models Faster (LIVE EVENT)
February 25, 2021, Build and Choose Better Models Faster (ON-DEMAND ENTRY)
For starters, consider sharing the "New In JMP16" page. The new Model Screening feature in JMP 16 allows users to automatically fit multiple predictive models and help select the best performing model and fit all available predictive models for categorical or continuous responses, with default tuning parameters. Resources will continue to be added to this page (including a new 1-minute highlight video). Available now is this awesome blog from Mia Stephens - it goes into great detail about Model Screening in JMP Pro 16.
Other resources include:
Build Better Models – Five Part Webinar Series
Though a fairly large commitment because it has so much content, this five-part webinar series is a good follow up for Kemal’s content. The series shows you how to build better and more useful models using modern predictive modeling techniques such as regression, neural networks and decision trees. The landing page can be reached by clicking on the link above. The individual webinars can be accessed using these links. (Most include case studies.)
Understanding the Principles for Building Predictive Models
This webinar helps attendees understand the foundation of effective model building. Covers a number of topics including Linear Regression, Stepwise Regression, Generalized Regression, Penalized Regression, Decision Trees, Bootstrap Forests, Boosted Trees and Neural Networks. Also includes discussions around the importance of test sets and how to use them and comparing and choosing models. (The webinar uses JMP 14 and JMP Pro 14.)
Building Better Predictive Models Using JMP Pro
This August 2020 webinar featuring Andrea Coombs demonstrates JMP Pro’s predictive modeling capabilities and basic predictive model types and helps users understand suggested workflow and which model to use at each step. The webinar includes a case study (Diabetes Disease Progression). Also includes discussions on variable selection, building Generalized Regression model using Adaptive Lasso estimation method and building multiple candidate prediction models (Bootstrap Forest and Boosted Tree) that include risk factors identified during variable selection. Concludes with more than 5-minutes of Q&A.
Solving science and engineering problems with machine learning, Stat Speaking Event
Very strong resource, high-quality resource. This was a Stat Speaking, Europe event, so it includes European-based speakers, but in-depth discussion (keynote) on deep learning, AI with Dick DeVeaux and a strong/engaging panel.
What is Machine Learning (and can JMP do it), JMP Blog
1st part of Elizabeth Claassen’s blog series. Good introductory discussion of concepts around machine learning and outlines all the different machine learning algorithms JMP can do. Promotes the free trial.
Data Mining and Predictive Modeling, JMP Tutorial
This JMP Tutorial includes a manufacturing case study and includes demonstrations of a number of techniques using Graph Builder and the Prediction Profiler and JMP Pro. Discussions on topics related to machine learning and AI, including regression, Bootstrap Forest and Boosted Trees (Decision Trees) and Neural Networks.
Building Better Models Using JMP Pro, by Jim Grayson, Mia L. Stephens and Sam Gardner
Recommending two resources here.
- The first is a free book chapter posted to our JMP Resource Center. Chapter 7 from Building Better Models With JMP® Pro, demonstrates the “what, why and how” of neural networks through case studies, figures and exercises. This book is designed for business analysts, managers, and practitioners who may not have a solid statistical background but need to be able to readily apply analytic methods to solve business problems.
- The second is the full book from SAS Press, which can be bought from SAS Press here.
Additionally, these sessions include discussions on tree-based models and neural networks - you may want to consider sharing these as well:
Understanding and Applying Tree-Based Methods for Predictor Screening and Modeling, Mastering JMP Webinar
Discovering and Predicting Patterns Using Neural Network Models, Mastering JMP Webinar
