Fast algorithms for minimum homology basis

From MaRDI portal





This works describes three new randomized algorithms for calculating an optimal, i.e., a \textit{minimum} homology basis of the 1-dimensional homology group (given \(\mathbb{Z}_2\) coefficients). With all three algorithms have different computational complexities, the third one notably achieves nearly quadratic runtime in case the homology group has constant rank. This is result, while being of independent interest in computational topology, also lends itself to various applications like network analysis. Next to providing a detailed proof of the correctness and computational complexity of the described algorithms, the paper also contains pseudocode description of all algorithms as well as a discussion on how to implement the moist suitable algorithm in practice. This is complemented by an analysis of empirical performance on real-world and synthetic data, demonstrating the benefits of the proposed method.











This page was built for publication: Fast algorithms for minimum homology basis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6963486)