New to Gaussian Processes question on single output and multiple inputs

So please forgive my ignorance for this question but can a GP with a single output variable I want to predict take multiple input variables?

If so, does anyone have an example of this being implemented?

Also, if the inputs might interact with each other in some way, are there any special things I need to worry about beyond the normal things one idea with an OLS for example?

Absolutely. You can specify the input_dim for the covariance function to be 2 and it will expect two columns of input. Here’s an example notebook.

1 Like

Thank you and you are my favorite presenter on PYMC. :slight_smile:

Hey Thanks! @bwengals gave a really nice talk on GPs in PyMC recently too/