Level: Intermediate
【Presenter】
Mia Stephens, JMP Principal Product Manager, SAS Institute Inc.
[Announcement summary]
Predictive modeling is finding the model or combination that most accurately predicts the outcome of interest. However, not all problems (or data) occur in the same way. You can fit several predictive models in any scenario, and on the other hand, there is not one model that can solve all problems. Regression models may perform best, while tree models and neural networks may perform best.
When looking for the best performing model, you can use cross-validation to fit all available models at once. You can then save the individual models in a data table or formula depot and use Model Comparison to compare the performance of the models in the validation set and select the best model. The new model screening platform included in JMP Pro 16 has streamlined this workflow. In this presentation, you will learn how to use model screening to simultaneously perform fit, validation, comparison, exploration, and selection to introduce an optimal predictive model.
* Play the video on this site, click "CC" at the bottom right of the screen, and select "English" to display English subtitles.
If you would like to download the presentation materials and data of this announcement, or watch them together with the Japanese transcript, Here Please click on the.
Go to the page for each presentation in the Discovery Summit Americas 2021 Presentation.
(Please note that you will be asked to log in to SAS Profile.)
