Skip to content

Latest commit

 

History

History
17 lines (11 loc) · 951 Bytes

README.md

File metadata and controls

17 lines (11 loc) · 951 Bytes

Introduction to Kernel Ridge Regression

place to kee notes, code, plots and pdf for my final project in kernel ridge regression

About:

This is a repo for my final project in Advanced Statistical Methods. The project intends to

  • introduce non-linear regression
  • explain Repreoducinng Kernel Hilbert Spaces, basis expansions, and its relationship to Kernel Ridge Regression
  • explore a few simulated data sets with KRR
  • disuss/execute parameter tuning in the context of generalizability and cross-validation.
  • mention a few contexts in which KRR is used and discuss comparable models

Code:

Two primary packages are used: CVST and DRR. CVST allows for cross-validation via sequential testing. DRR allows for faster KRR via dimension reduction using a generalized PCA approach. Simulate data from CVST. Choose parameters: lambda and sigma.