Decorrelation Overview

Decorrelation uses linear correlation techniques to allow for the identification of inputs that are highly correlated.  Decorrelating the data set by removing (or "marking") inputs that are highly correlated from the selected model inputs will result in a more reliable model of the process, reducing the number of inputs to the model, and thereby reducing the complexity of the model.  

Show me how >>

Related topics:

• Decorrelating the dataset

• Setting process limits

• Data visualization: Trends

• Data visualization: Scatter plots

• Data visualization: Statistics

• Data visualization: Regional brushing

• Visualization Overview