This website uses cookies to track visitor behaviour anonymously - this helps us to improve the website.
Read our Privacy Policy for more information.    Accept | Do Not Track
Home » Knowledge » Techniques » Principal Component Regression

Principal Component Regression

Principal Component Regression (PCR) is a regression method that uses PCA initially to reduce a large data set (X) into a relatively small number of principal components, and then performs a linear regression (MLR) between these components and sample properties (y). It is particularly useful for analysing spectroscopic data, in which the number of variables is often significantly larger than the number of samples, which prevents an MLR method to be used on the raw data. Typically a PCR method would use some kind of validation method (e.g. cross-validation) to ensure robustness.