Regression using RBF kernel

[Nadheesh]

@junpenglao

No, you need to conditioned one the test set to make prediction on the test set. In another word, you need to reevaluate the cov (and mean) function on the X_test to make prediction of y_test

I don’t understand how to implement this. Can’t we directly use the GP module. It says that gp.Marginal can be used for regression with normally distributed data. Is not it similar to RBF?

I used this code to train the model as shown in the example :

with pm.Model() as model:
    # Specify the covariance function.
    cov_func = pm.gp.cov.ExpQuad(1, ls=0.1)

    # Specify the GP.  The default mean function is `Zero`.
    gp = pm.gp.Marginal(cov_func=cov_func)

    # Place a GP prior over the function f.
    sigma = pm.HalfCauchy("sigma", beta=3)
    y_ = gp.marginal_likelihood("y", X=X, y=y, noise=sigma)

Then used gp.predict(X_new) to predict. Can you please provide me some more direction?

Thanks.