Skip to main content
Webinars

[On Demand] Mitigating Prediction Errors in Empirical Models​ with SVEM and SFD

Mitigate Predictive Error with Predictum’s SVEM Add-In for JMP

Click the image above for on-demand access to the webinar!

In science and engineering, empirical models built with Design of Experiments and regression analysis have long helped experts predict performance in multivariate systems. Yet, when applied to future predictions, these models sometimes fall short. Shifts in the process, large prediction errors, or bias in the model can all cause results to miss the mark.
 
Today, new methods in machine learning for small experimental data offer practical ways to mitigate these issues and achieve more reliable predictions. By combining Self-Validating Ensemble Models (SVEM) with Space Filling Designs, it’s possible to reduce bias and precision errors and strengthen the accuracy of empirical models.
 
In this webcast, Predictum’s Cy Wegman, will demonstrate how science and engineering professionals can use SVEM and Space Filling Designs to obtain higher accuracy predictions than one might find via traditional approaches.

In this session, you will:

  • Understand how SVEM and Space Filling Designs can mitigate prediction errors in empirical models
  • Learn to build accurate, reliable predictive models that can handle complex science and engineering systems using the SVEM Methodology
  • Identify why traditional data collection and modeling methods fail to accurately predict for complex systems​​

Presenters:

Cy Wegman
Senior Analyst

Cy wegman

Cy specializes in data analytics, empirical modeling and optimization.

Cy Wegman is a Senior Analyst at Predictum. He specializes in data analytics and empirical modeling and optimization. Before joining Predictum, he worked for 38 years at Procter & Gamble, where his last assignment was the Empirical Modeling leader for the company. Cy is a member of the P&G Prism Society, which represents the top 1% of P&G engineers, for profitably applying his technical mastery. His contributions have resulted in hundreds of millions of dollars in annual savings. His cross-industry work in manufacturing, engineering, and R&D has spanned molecular-scale to full-scale technologies, as well as material, formulation, process, packaging, and consumer models. Cy holds a B.S. in Civil Engineering from Rose-Hulman Institute of Technology.

Webinar resources and contact info:

EPISODE LINKS:
Cy Wegman’s work profile: https://predictum.com/company/#leadership
Predictum SVEM Product Page: SVEM
SVEM Frequently Asked Questions: Frequently Asked Questions about SVEM
Free SVEM Micro-Course: https://courses.predictum.com/courses/build-better-and-faster-with-small-experimental-designs-and-machine-learning

PEER-REVIEWED RESEARCH ON SVEM:

CONTACT:
Predictum Inc – Contact Predictum

SHARE FEEDBACK ON THIS WEBINAR:
Submit Feedback – Feedback via Microsoft Forms

SOCIAL LINKS:
LinkedIn (Cy Wegman): https://linkedin.com/in/cywegman
LinkedIn (Nathaniel Leies): https://linkedin.com/in/nathanielleies
LinkedIn (Predictum Inc.): https://linkedin.com/company/predictum

Try Predictum SVEM for Free (30-Day Evaluation License)

Requires active JMP License. Try JMP for free at jmp.com.

     

    Predictum Inc. is committed to protecting your personal information. The information that you provide to us will only be used in accordance with the applicable data privacy laws, our Privacy Policy, our Terms of Use, and our internal policies.

     

     

    Mockup of the Predictum Insider Newsletter.

    1,000+ scientists, engineers, and analysts read our monthly briefing to scale analytical programs in their organizations

    Get the latest methods, tools, and use cases delivered to your inbox