spectral representation of graphon convolutions

[[concept]]

Theorem

Given a graphon convolution with input graphon signal and output , the WFTs of are related as

{\cal \hat{Y}}_{j} &= \sum_{k=0}^{K-1} h_{k} \lambda_{j}^k {\cal \hat{X}}_{j} \\ &=\hat{T}_{H}(\lambda_{j}) . {\cal \hat{X}}_{j} \\ &= h(\lambda_{j}) {\cal \hat{X}}_{j} \end{align}$$ Where $h(\lambda_{j})=\sum_{k=0}^{K-1} h_{k} \lambda_{j}^k$ is our polynomial. Here, $\hat{T}_{H}$ acts *pointwise* on the graphon signal ${\cal X}$

NOTE

Proof

{\cal Y} &= \sum_{k=0}^{K-1} h_{k} \sum_{j \in \mathbb{Z} \setminus \{ 0 \}} \lambda_{j}^k \varphi_{j}(v) \underbrace{ \int _{0}^1 \varphi_{j}(u) {\cal X}(u) \, du }_{ \hat{x}_{j} } \\ &= \sum_{k=0}^{K-1} h_{k} \sum_{j \in \mathbb{Z} \setminus \{ 0 \}} \lambda_{j}^k \varphi_{j}(v) {\cal \hat{X}} \\ \implies \hat{{\cal Y}}_{i} &= \int _{0}^1 {\cal Y}(u) \varphi_{i}(u) \, du \\ &= \sum_{k=0}^{K-1} h_{k}\lambda_{i}^k \cancelto{1}{ \int \varphi_{i}^2(u)\, du } \hat{x}_{i} \\ \implies \hat{{\cal Y}}_{i} &= \sum_{k=0}^{K-1}h_{k}\lambda_{i}^k \hat{x}_{i} \end{align}$$ This is a simple result following the [[Concept Wiki/diagonalizable\|diagonalization]] of $T_{W}^{(k)}$ by $\varphi_{i}$

Note

Some interesting takeaways (analogues of the graph convolution)

  • Like the WFT, the spectral response of a graphon convolution is discrete
  • the th spectral component of the output only depends on and spectral graphon signal
    • ie depends only on and
  • The spectral response of given by is independent of the underlying (like the spectral response of was independent of the graph)
  • ie, the spectral response can be written as a polynomial function. The eigenvalues of the graphon determine where we evaluate this function. (samples of function that is eigenvalues/spectrum)

Example

Sampling eigenvalues from the spectrum polynomial

fixed coefficients yield the same spectral response for both graphon and graph convolutions

Important

Given the same (fixed) coefficients , define the graph convolution . The spectral representation of that convolution is

Review

dsg

In the spectral domain, the graphon shift operator {acts pointwise||special behavior} on the graphon signal

Like the GFT, the spectral response of a graphon convolution is {discrete||property}

The th spectral component of the output only depends on eigenvalue and the spectral graph signal

The spectral response of graphon convolution given by is {==independent of the underlying graphon ||relation to object==}

the spectral response can be written as {1||a polynomial||function}. The eigenvalues of the {2||graphon shift operator} determine {3||where we evaluate} this function.

Mentions

Mentions

TABLE file.mday as "Last Modified"
FROM [[]]
 
FLATTEN choice(contains(artist, this.file.link), 1, "") + choice(contains(author, this.file.link), 1, "") + choice(contains(director, this.file.link), 1, "") + choice(contains(source, this.file.link), 1, "") as direct_source
 
WHERE !direct_source
SORT file.mday ASC
SORT file.name ASC
const { dateTime } = await cJS()
 
return function View() {
	const file = dc.useCurrentFile();
	return <p class="dv-modified">Created {dateTime.getCreated(file)}     ֍     Last Modified {dateTime.getLastMod(file)}</p>
}