Corpus ID: 214794874

Orthogonal Inductive Matrix Completion

@article{Ledent2020OrthogonalIM,
  title={Orthogonal Inductive Matrix Completion},
  author={Antoine Ledent and R. Alves and M. Kloft},
  journal={ArXiv},
  year={2020},
  volume={abs/2004.01653}
}
  • Antoine Ledent, R. Alves, M. Kloft
  • Published 2020
  • Mathematics, Computer Science
  • ArXiv
  • We propose orthogonal inductive matrix completion (OMIC), an interpretable model composed of a sum of matrix completion terms, each with orthonormal side information. We can inject prior knowledge about the eigenvectors of the ground truth matrix, whilst maintaining the representation capability of the model. We present a provably converging algorithm that optimizes all components of the model simultaneously, using nuclear-norm regularisation. Our method is backed up by \textit{distribution… CONTINUE READING

    Figures, Tables, and Topics from this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 90 REFERENCES

    Provable Non-linear Inductive Matrix Completion

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL

    Spectral Regularization Algorithms for Learning Large Incomplete Matrices

    VIEW 10 EXCERPTS
    HIGHLY INFLUENTIAL

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

    VIEW 3 EXCERPTS
    HIGHLY INFLUENTIAL

    Exact matrix completion via convex optimization

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Rademacher and Gaussian Complexities: Risk Bounds and Structural Results

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    The Power of Convex Relaxation: Near-Optimal Matrix Completion

    • E. Candès, T. Tao
    • Computer Science, Mathematics
    • IEEE Transactions on Information Theory
    • 2010
    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    A Comprehensive Survey on Graph Neural Networks