Using Prediction Profiling to Maximize Model Proficiency
Statistics, Predictive Modeling and Data Mining
Prediction profiling lets you interactively examine the impact on your response when you change individual factor settings while the others are held constant. See how to extend your proficiency to use the Profiler, available from 20+ JMP analysis platforms, to understand your models, set desirability goals for your response(s), find optimal factor settings, independently assess factor importance relative to model predictions, and simulate the distribution of model outputs as a function of the random variation in model inputs and noise.
This session covers: customizing Profiler appearance, desirability, optimization, simulation, error propagation, variable importance, and sensitivity indicator.