Choose Language Hide Translation Bar

Applications of Dynamic Regression Models in Business and Industry (2019-US-45MP-268)

Level: Intermediate

 

Bob Lucas, Principal, Robert M. Lucas Consulting

 

Dynamic regression models extend multiple regression models by allowing for independent variables to be incorporated as leading indicators of the dependent variable and to account for autocorrelation of the dependent variable. Dynamic regression models can be built in the JMP Time Series platform.

 

Manufacturing processes data is often a time series with process input characteristics and the final output quality varying over time. A dynamic regression model may be used to ascertain how to adjust process parameters to control the variability of process output quality.

In business, market mix models are used to evaluate the return on investment of different advertising strategies. A dynamic regression model can be used to evaluate the return on investment of advertising spend by different media, a market mix model. One can use the model to drive sales by allocating advertising budgets efficiently. 

 

The author will use the above examples to illustrate how to build dynamic regression models using JMP and to interpret report results.

 

 

Comments

Hello @robertlucas1972, Wonderful job!  (with both the presentation and your well-documented manuscript)

 

We recently got some general interest from our customer base about setting up and running dynamic linear regression (DLR) models in JMP, and your work is a wonderful introduction to the topic, in a very practical context.  Would it be possible for you to share the sample data tables that you demoed with here?  I think they would serve as a useful self-teaching tool for using the Time Series Platform in this context (with your manuscript to follow along).

 

Cheers, 

@PatrickGiuliano