Resources

Pairwise Distances

Libraries

Gaussian Processes

• LADAX

  GPs and DeepGPs. Using Layers of distributions.

class SVGP(nn.Module):
    def apply(self, x):
        kernel_fn = kernel_provider(x, **kernel_fn_kwargs)
        inducing_var = inducing_variable_provider(x, kernel_fn, **inducing_var_kwargs)
        vgp = SVGPLayer(x, mean_fn, kernel_fn, inducing_var)
        return vgp

MonteCarlo

• Numpyro
• MCX

Kernel Methods

• Kernel Density Estimation - jax_cosmo
• Gaussian Kernel

Polynomial

• Interpolation - jax_cosmo

Normalizing Flows

• Normalizing Flows in JAX - ChrisWaites

Statistical Learning

• Rethinking Statistical Learning - Fehlepsi