There are tons of books and a huge literature on the properties of the following integral operator: \begin{equation} T(f) = \int_{\mathcal{X}} K(x,\cdot)f(x)dx, \end{equation} where $K(x,z)$ is, say, a Mercer kernel.
I am wondering if people have systematically studied the properties (spectrum for example) of the following weighted integral operator: \begin{equation} T^*(f) = \int_{\mathcal{X}} K(x,\cdot)f(x)d\mu(x), \end{equation} where $\mu$ is a non-Lebesgue measure on $\mathcal{X}$. And I am imagining that if $\mu$ is closer to the Lebesgue measure, $T$ and $T^*$ should be more similar to each other. Are you all aware of anything on this topic? Thanks!
Here's an important theorem which relates kernels to Hilbert-Schmidt operators:
Hilbert-Schmidt operators are, in particular, compact operators. If $K(x,y)=\overline{K(y,x)},$ then $T$ is both compact and self-adjoint, which gives you nice spectral-theoretic results.