Predictive Modeling Using JMP® Pro

This course covers the skills required to develop, assess, and score predictive models using the Partition platform and the Neural platform. You will learn how to explore data graphically and how to build and understand predictive models such as decision trees and neural network models. You will compare and explain complex models and score new data in JMP and generate SAS score code.

Voraussetzungen
Before attending this course, it is recommended that you complete the courses „JMP Software: statistische Datenanalyse" (JDEX10) and „JMP Software: Varianzanalyse und Regression" (JANR10) or have equivalent experience.

Zielgruppe
Data analysts, qualitative experts, and others who want an introduction to predictive modeling using JMP Pro

Module
JMP Software

Kursinhalte

• Introduction to Predictive Modeling
  • supervised classification
  • regression
  • honest assessment using holdout data
  • assessment plots and statistics
• Exploring Data
• Predictive Modeling with the Partition Platform
  • recursive partitioning
  • optimizing model complexity
  • model fit statistics
  • statistical graphics
• Predictive Modeling with Neural Networks
  • optimizing model complexity
  • interpreting models
  • model fit statistics
  • statistical graphics
  • k-fold crossvalidation
  • boosted neural networks
• Tree Advanced Topics
  • instability of trees
  • boosted trees
  • bootstrap forest
  • comparing models

zurück...