• Corpus ID: 227275676

Matrix Completion with Hierarchical Graph Side Information

@article{Elmahdy2020MatrixCW,
  title={Matrix Completion with Hierarchical Graph Side Information},
  author={Adel M. Elmahdy and Junhyung Ahn and Changho Suh and Soheil Mohajer},
  journal={ArXiv},
  year={2020},
  volume={abs/2201.01728}
}
We consider a matrix completion problem that exploits social or item similarity graphs as side information. We develop a universal, parameter-free, and computationally efficient algorithm that starts with hierarchical graph clustering and then iteratively refines estimates both on graph clustering and matrix ratings. Under a hierarchical stochastic block model that well respects practically-relevant social graphs and a low-rank rating matrix model (to be detailed), we demonstrate that our… 

Figures and Tables from this paper

When to Use Graph Side Information in Matrix Completion

TLDR
The key idea is to make a careful selection for the information employed in the first clustering step, between two types of given information: graph & matrix ratings.

The Optimal Sample Complexity of Matrix Completion with Hierarchical Similarity Graphs

TLDR
The information-theoretic limit on the number of observed matrix entries is characterized as a function of the quality of graph side information by proving sharp upper and lower bounds on the sample complexity.

On the Fundamental Limits of Matrix Completion: Leveraging Hierarchical Similarity Graphs

TLDR
The optimal sample complexity is analyzed and different regimes whose characteristics rely on quality metrics of side information of the hierarchical similarity graph are identified and it is shown that the characterized information-theoretic limit can be asymptotically achieved.

Graph-assisted Matrix Completion in a Multi-clustered Graph Model

  • Geewon SuhChangho Suh
  • Computer Science
    2022 IEEE International Symposium on Information Theory (ISIT)
  • 2022
TLDR
A computationally efficient algorithm is developed that achieves the optimal sample complexity for the entire regime of graph information under the multiple cluster setting and outperforms prior algorithms that leverage graph side information.

Community Detection and Matrix Completion With Social and Item Similarity Graphs

TLDR
This work considers the problem of recovering a binary rating matrix as well as clusters of users and items based on a partially observed matrix together with side-information in the form of social and item similarity graphs together with lower and upper bounds on sample complexity that match for various scenarios.

Discrete-Valued Latent Preference Matrix Estimation with Graph Side Information

TLDR
A new model is proposed in which 1) the unknown latent preference matrix can have any discrete values, and 2) users can be clustered into multiple clusters, thereby relaxing the assumptions made in prior work.

Online Low Rank Matrix Completion

TLDR
A novel algorithm OCTAL (Online Collaborative filTering using iterAtive user cLustering) that ensures nearly optimal regret bound of O ( polylog ( M + N ) T 1 / 2 ) .

Nonparametric Matrix Estimation with One-Sided Covariates

  • C. Yu
  • Computer Science
    2022 IEEE International Symposium on Information Theory (ISIT)
  • 2022
TLDR
This work provides an algorithm and accompanying analysis which shows that the algorithm improves upon naively estimating each row separately when the number of rows is not too small and achieves the minimax optimal nonparametric rate of an oracle algorithm that knows the row covariates.

References

SHOWING 1-10 OF 87 REFERENCES

On the Fundamental Limits of Matrix Completion: Leveraging Hierarchical Similarity Graphs

TLDR
The optimal sample complexity is analyzed and different regimes whose characteristics rely on quality metrics of side information of the hierarchical similarity graph are identified and it is shown that the characterized information-theoretic limit can be asymptotically achieved.

Matrix Completion on Graphs

TLDR
This work introduces a novel matrix completion model that makes use of proximity information about rows and columns by assuming they form communities, and borrows ideas from manifold learning to constrain the solution to be smooth on these graphs, in order to implicitly force row and column proximities.

Achievability Bounds for Community Detection and Matrix Completion with Two-Sided Graph Side-Information

TLDR
A lower bound on the minimum expected number of observed ratings (also known as the sample complexity) needed for this recovery task is developed, which is a function of various parameters including the quality of the graph side-information manifested in the intra-and inter-cluster probabilities of the SBMs.

Binary Rating Estimation with Graph Side Information

TLDR
To the best of the knowledge, this work is the first to reveal how much the graph side information reduces sample complexity, and proposes a computationally efficient algorithm that achieves the limit.

Kernelized Probabilistic Matrix Factorization: Exploiting Graphs and Side Information

TLDR
A new matrix completion algorithm— Kernelized Probabilistic Matrix Factorization (KPMF), which effectively incorporates external side information into the matrix factorization process, and can be easily used with other types of side information as well.

Collaborative Filtering with Graph Information: Consistency and Scalable Methods

TLDR
This work formulate and derive a highly efficient, conjugate gradient based alternating minimization scheme that solves optimizations with over 55 million observations up to 2 orders of magnitude faster than state-of-the-art (stochastic) gradient-descent based methods.

Community Detection and Classification in Hierarchical Stochastic Blockmodels

TLDR
This work proposes a robust, scalable, integrated methodology for community detection and community comparison in graphs, and addresses the problem of locating similar sub-communities in a partially reconstructed Drosophila connectome and in the social network Friendster.

Robust online matrix completion on graphs

TLDR
This work proposes an algorithm that exploits the graph to reconstruct the incomplete data, in the scenario where there is outlier noise, and demonstrates the improved performance of the proposed technique compared to other state of the art algorithms.

Query Complexity of Clustering with Side Information

TLDR
The dramatic power of side information aka similarity matrix on reducing the query complexity of clustering is shown, and intriguing connection to popular community detection models such as the {\em stochastic block model}, significantly generalizes them, and opens up many venues for interesting future research.

Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks

TLDR
This paper proposes a novel approach to overcome limitations of matrix completion techniques by using geometric deep learning on graphs, and applies this method on both synthetic and real datasets, showing that it outperforms state-of-the-art techniques.
...