Topological data analysis is a recent and fast growing field that approaches the analysis of datasets using techniques from (algebraic) topology. Its main tool, persistent homology (PH), has seen a notable increase in applications in the last decade. Often cited as the most favourable property of PH and the main reason for practical success are the stability theorems that give theoretical results about noise robustness, since real data is typically contaminated with noise or measurement errors. However, little attention has been paid to what these stability theorems mean in practice. To gain some insight into this question, we evaluate the noise robustness of PH on the MNIST dataset of greyscale images. More precisely, we investigate to what extent PH changes under typical forms of image noise, and quantify the loss of performance in classifying the MNIST handwritten digits when noise is added to the data. The results show that the sensitivity to noise of PH is influenced by the choice of filtrations and persistence signatures (respectively the input and output of PH), and in particular, that PH features are often not robust to noise in a classification task.
Computation of degree-1 persistent homology on larger point-clouds using the Reduced Vietoris-Rips filtration
Australian National University(澳大利亚国立大学);Centre for Systems Biology Dresden(德累斯顿系统生物学中心);Max Planck Institute of Cell Biology and Genetics(马克斯·普朗克细胞生物学与遗传学研究所);Max Planck institute for the Physics of Complex Systems(马克斯·普朗克复杂系统物理研究所)
Computing extended persistent homology of radial distance filtrations of Euclidean shapes
Division of Applied Mathematics, Brown University(布朗大学应用数学系);Research School of Physics, Australian National University(澳大利亚国立大学物理研究学院);Mathematical Sciences Institute, Australian National University(澳大利亚国立大学数学科学研究所)
EPFL(洛桑联邦理工学院);IST Austria (Institute of Science and Technology Austria)(奥地利科学技术研究所);Mathematical Sciences Institute, The Australian National University(澳大利亚国立大学数学科学研究所);University of New England(新英格兰大学);Research School of Physics, The Australian National University(澳大利亚国立大学物理研究学院)
Parallel multi-scale reduction of persistent homology filtrations
持久同源管道包括所谓的边界矩阵的简化。我们扩展了Bauer et al.(2014)和Chen et al.(2011)的工作,他们展示了如何使用边界矩阵中的依赖关系来适应Edelsbrunner et al.(2002)中提出的约简算法,以降低其计算成本。在这里,我们提出了一些额外的依赖关系的边界矩阵,并提出了一种新的并行算法减少边界矩阵。特别是,我们显示:这部分的减少是立即明显的,给出了对剩余列所需的减少的界限,并从这些给出了一个框架,对于该框架,边界减少过程可以大规模并行化。四个合成的例子的模拟表明,计算负担可以进行约千分之一的迭代次数所需的传统方法。此外,尽管传统的边界简化从过滤顺序顺序地揭示条形码,但是该方法给出了一种替代方法,通过该方法,条形码同时针对多个尺度被部分地揭示,并且随着算法的进展而被进一步细化;模拟显示对于具有$\sim10^4$单纯形的Vietoris-Rips滤波,在两次迭代中可以计算出具有95%精度的基本单形的估计,并且约简完成在1%以内在我们的算法的大约10次迭代中,与传统方法的大约8000次迭代相反。
Motif-based filtrations for persistent homology: A framework for graph isomorphism and property prediction
Center for Complex Networks and Systems Research, Luddy School of Informatics, Computing, and Engineering, Indiana University(印第安纳大学卢迪信息、计算与工程学院复杂网络与系统研究中心);Delft University of Technology(代尔夫特理工大学);Universitat de Barcelona(巴塞罗那大学);Institute of Complex Systems (UBICS), Universitat de Barcelona(巴塞罗那大学复杂系统研究所);ICREA(加泰罗尼亚高级研究院)
Mix-GENEO: A Flexible Filtration for Multiparameter Persistent Homology Detects Digital Images
School of Mathematics, Jilin University(吉林大学数学学院);School of Artificial Intelligence, Jilin University(吉林大学人工智能学院);School of Mathematical Science, Heilongjiang University(黑龙江大学数学科学学院)
Persistence Pairing Graphs: Tracking Changes of Selected Homology Bases
Dioscuri Centre in Topological Data Analysis, Institute of Mathematics of the Polish Academy of Sciences(波兰科学院数学研究所拓扑数据分析迪奥斯库里中心);International Environmental Doctoral School, University of Silesia in Katowice(卡托维兹西里西亚大学国际环境博士学院)
本文发展了加权持久同调理论。1990年,Robert J. Dawson首先深入研究了加权单纯复形的同调性。我们推广了加权单纯复形的定义和加权单纯复形的同调性,使权可以在整环$R$中存在。然后我们研究由此产生的加权持久同调。我们表明,加权持久同源性可以区分过滤,普通的持久同源性不区分。例如,如果有一个点被认为是特殊的,加权持久同调可以告诉包含该点的循环何时形成或消失。