Photo by Ales Krivec on Unsplash

by Yubo 'Paul' Yang

Kriging

Kriging is an interpolation method that provides confidence interval to its predictions. Kriging works by minimizing the variance of the prediction error over existing data.

Presentation Summary

In these slides, I present:

  • Basic idea behind the kriging algorithm.
  • Implementation of ordinary kriging applied to 1D interpolation.

Examples

  • okrig.py: ordinary kriging on noisy sin(x) data
  • skl.py: sklearn gpr applied to data imported from okrig.py

References

All Signal Processing

ALGORITHM
signal processing machine learning