Timeline for Inverse transform of $\int K(s,t) \rho(s) dS $
Current License: CC BY-SA 4.0
4 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Mar 30, 2022 at 13:46 | comment | added | Dirk | A compact operator on a Hilbert space has a singular value decomposition which can be obtained by application of the spectral theorem for compact and self-adjoint operators to the operator $F^*F$. This may be found in books on functional analysis (while most treat the spectral theorem, not all treat the singular value decomposition). A domain in R^d is just an open and connected subset of R^d. It may be that everything is the same if $V$ is a surface, but I haven't thought about it. | |
Mar 30, 2022 at 10:11 | comment | added | MikeTeX | Note: I'm not sure what you mean by "domain in $\mathbb R^d$". In my setting, $V$ is a surface in $\mathbb R^d$ or the union of disjoint surfaces. I meant a differentiable manifold of dim d-1, or the union of such manifolds. To fix my ideas, I was considering objects like spheres, ellipsoids etc. in $\mathbb R^3$. | |
Mar 30, 2022 at 10:02 | comment | added | MikeTeX | Thank you for answering me. My knowledge is limited in this domain, and I lost you when you said "Consequently, the singular values of F go to zero and hence, no continuous inverse exist". Could you elaborate a bit in your answer? Regarding my motivating pseudo-demonstration, I am of course aware that it is not viable, even as an "approach" to this problem; it was just "kind of" informal approach. | |
Mar 30, 2022 at 9:24 | history | answered | Dirk | CC BY-SA 4.0 |