Subscribe Bookmark RSS Feed

Community Manager


Feb 27, 2013

Finding a good regression model with JMP

Systems engineer Jami Hampton opened the autumn 2011 series of live Mastering JMP webcasts to 600 JMP users who wanted to learn how to find a good multiple regression model using the JMP Fit Model platform.

Jami's PowerPoint presentation and videos of her demo are available.

Jami used data and a business case study from the book JMP Means Business – Statistical Models for Management by Josef Schmee and Jane Oppenlander. Thanks go to the authors for granting us permission to use the case study and for allowing us to distribute the data, with scripts added by Jami, so you can try building the regression model yourself.

Watch the videos and see for yourself how she:

  • Determined that deposits, total staff, loan sales and transactions are drivers.
  • Determined loan sales and transactions are correlated, so loan sales could be removed from the model.
  • Found a logical effective model that was easy to explain.
  • Jami's systems engineer colleagues are presenting Mastering JMP sessions Thursdays through Nov. 17. Join them as they show you how to delve into JMP and expand your JMP skills. Registration is required.

    1 Comment
    Community Member

    Daser Sunday wrote:

    i will recommend you to my blog to read about it.

    I have written about Linear Regression before using gradient descent. I figure it can be of help. However i concentrated my writeup on uni-variate regression only. That is when you are combated with just a single training set (X and Y). It is a great read though. I will import it soon on stack overflow.