Spectral Lifting (Graph to Hypergraph)

[AS-L-C]

This PR implements a new Graph :arrow_right: Hypergraph lifting method inspired by the spectral clustering algorithm proposed by Ng, Jordan, and Weiss (2002) [1], which leverages the graph's spectral properties.

We provide a formal definition of the spectral lifting method below, following the notation introduced by von Luxburg in [2].

---

Algorithm: Spectral Lifting (Graph :arrow_right: Hypergraph)

---

Input: Graph (weighted) adjacency matrix $W \in \mathbb{R}^{n \times n}$, where $n$ is the number of nodes

• Compute the Laplacian matrix $L = D - W$, where D is the degree matrix.
• Compute the normalized Laplacian $L_{\text{sym}} = D^{-1/2}LD^{-1/2}$.
• Compute the first $k$ eigenvectors $u_1, \ldots, uk$ of $L{\text{sym}}$.
• Let $U \in \mathbb{R}^{n \times k}$ be the matrix containing the vectors $u_1, \ldots, u_k$ as columns.
• Form the matrix $T \in \mathbb{R}^{n \times k}$ from $U$ by normalizing the rows to norm 1, that is set $t{ij} = \frac{u{ij}}{(\sum{c} u{ic}^2)^{1/2}}$.
• For $i = 1, \ldots, n$, let $y_i \in \mathbb{R}^k$ be the vector corresponding to the $i$-th row of $T$.
• Cluster the points $(yi){i=1, \ldots, n}$ into clusters $C_1, \ldots, C_k$ with a soft or hard clustering algorithm (e.g., k-means).
• For $i=1,\ldots,k$ define $H_i$ as the hyperedge connecting all the points belonging to cluster $C_i$
• Build the hypergraph incidence matrix $M$ of the hypergraph ${\it H}:=\{Hi\}{{i=1, \ldots, k}}$.

Output: Hypergraph incidence matrix $M \in \mathbb{R}^{n \times k}$, where $k$ is the number of hyperedges

---

Notes

Note₁: the number of hyperedges $k$ can be automatically determined by using a heuristic based on the eigengaps (i.e., spectral gaps), which estimates the number of connected components in the original graph, or provided by the user.

Note₂: hyperedges with overlapping hypernodes can be obtained by using soft clustering methods, while some degree of randomness can be introduced by using non-deterministc clutering methods.

---

References

[1] Ng, Andrew, Michael Jordan, and Yair Weiss. "On spectral clustering: Analysis and an algorithm." Advances in neural information processing systems 14 (2001). [2] Von Luxburg, Ulrike. "A tutorial on spectral clustering." Statistics and computing 17 (2007): 395-416.

---

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[gbg141]

Hello @AS-L-C! Thank you for your submission. As we near the end of the challenge, I am collecting participant info for the purpose of selecting and announcing winners. Please email me (or have one member of your team email me) at guillermo_bernardez@ucsb.edu so I can share access to the voting form. In your email, please include:

• your first and last name (as well as any other team members)
• the title of the method you implemented
• the input domain of the method you implemented
• the output domain of the method you implemented
• your pull request number (#68)

Before July 12, make sure that your submission respects all Submission Requirements laid out on the challenge page. Any submission that fails to meet this criteria will be automatically disqualified.