Using Partial Least Squares: When Ordinary Least Squares Regression Just Won’t Work

See how to:

• Understand when and why PLS is useful as well as the advantages of the PLS implementation in JMP Pro over that in JMP

See how to:

• Use consumer ratings to identify product attributes to help guide new formulation and design processes
  • Employ leave-one-out cross-validation,
  • Examine and interpret  Root Mean Press; NIPALS Fit x and y scores for a single latent factor; Diagnostic Plots; and VIP vs Coefficients Plots.
  • Use Prediction Profiler to maximize desirability.

See how to:

• Create a model to evaluate the levels of three different compounds in spectral emissions of water samples
  • Use Model Comparison summary to see how x and y scores correlate and to identify which variables explain variation
  • Save prediction values to the data table, to make the model useful if new observations are added

See how to:

• Use Discriminant Analysis in a dataset that includes over 10,000 gene expression characteristics from a study of 230 individuals
  • Determine if genetic expression information can be used to accurately classify estrogen receptor status
  • Explore missing values; interpret ROC curves; and examine misclassification, false positives and false negatives

Resources

• SAS Press Book: Discovering Partial Least Squares with JMP^®