Skip to main content
Ctrl+K

GPyTorch-QR

  • GPQR Tutoral
  • Examples
  • Module reference
  • GPQR Tutoral
  • Examples
  • Module reference

Section Navigation

  • Basic Usage
    • Direct GPQR (independent quantiles)
    • Center-gap GPQR (independent gaps)
    • Direct GPQR (correlated quantiles)
    • Center-gap GPQR (correlated gaps)
  • Placing the Prior Mean
    • Subtracting the mean
      • Direct GPQR
      • Center-gap GPQR
    • Setting the mean module
      • Direct GPQR (independent)
      • Direct GPQR (correlated)
      • Center-gap GPQR (independent)
      • Center-gap GPQR (correlated)
  • Multi-input GPQR
    • Center-gap GPQR
    • Scaling with Prior Mean
    • Length Scale Constraint
  • Multi-output GPQR
    • Direct GPQR
    • Center-Gap GPQR (independent centrals)
    • Center-Gap GPQR (correlated centrals)
  • Cross validation
    • Cross validation
  • Examples
  • Placing the Prior Mean

Placing the Prior Mean#

Prior mean function can be used to provide informative bias to quantiles, which is useful when the data is scarse.

There are two ways to place the prior mean \(\mu(x)\). The first is subtracting the prior mean from the input data \(y\). The second is setting the mean module to the prior distribution itself.

  • Subtracting the mean
    • Direct GPQR
    • Center-gap GPQR
  • Setting the mean module
    • Direct representation
    • Center-gap representation

previous

Center-gap GPQR (correlated gaps)

next

Subtracting the mean

Show Source

© Copyright 2026, Jisoo Song.

Created using Sphinx 9.1.0.

Built with the PyData Sphinx Theme 0.19.0.