### Abstract

widely used tool in this domain is persistent homology. By processing the data at all scales, it

does not rely on a particular choice of scale, which is one of the main challenge faced in this

area. Moreover, its stability properties provide a natural connection between discrete data

and an underlying continuous structure. Finally, it can be combined with other tools, like the

distance to a measure, which allows to handle noise that are unbounded. The main caveat of

this approach is its high complexity.

In this thesis, we will introduce topological data analysis and persistent homology, then show

how to use approximation to reduce the computational complexity. We provide an approxima-

tion scheme to the distance to a measure and a sparsifying method of weighted Vietoris-Rips

complexes in order to approximate persistence diagrams with practical complexity. We detail

the specific properties of these constructions.

Persistent homology was previously shown to be of use for scalar field analysis. We provide

a way to combine it with the distance to a measure in order to handle a wider class of noise,

especially data with unbounded errors. Finally, we discuss interesting opportunities opened

by these results to study data where parts are missing or erroneous.

Original language | English |
---|---|

Publication status | Published - 1 Dec 2014 |

Externally published | Yes |

### Fingerprint

### Cite this

**Topological Inference from Measures.** / Buchet, Mickael Laurent.

Research output: Thesis › Doctoral Thesis › Research

}

TY - THES

T1 - Topological Inference from Measures

AU - Buchet, Mickael Laurent

PY - 2014/12/1

Y1 - 2014/12/1

N2 - In this thesis, we study the problem of analysing topological structure in point cloud data. Onewidely used tool in this domain is persistent homology. By processing the data at all scales, itdoes not rely on a particular choice of scale, which is one of the main challenge faced in thisarea. Moreover, its stability properties provide a natural connection between discrete dataand an underlying continuous structure. Finally, it can be combined with other tools, like thedistance to a measure, which allows to handle noise that are unbounded. The main caveat ofthis approach is its high complexity.In this thesis, we will introduce topological data analysis and persistent homology, then showhow to use approximation to reduce the computational complexity. We provide an approxima-tion scheme to the distance to a measure and a sparsifying method of weighted Vietoris-Ripscomplexes in order to approximate persistence diagrams with practical complexity. We detailthe specific properties of these constructions.Persistent homology was previously shown to be of use for scalar field analysis. We providea way to combine it with the distance to a measure in order to handle a wider class of noise,especially data with unbounded errors. Finally, we discuss interesting opportunities openedby these results to study data where parts are missing or erroneous.

AB - In this thesis, we study the problem of analysing topological structure in point cloud data. Onewidely used tool in this domain is persistent homology. By processing the data at all scales, itdoes not rely on a particular choice of scale, which is one of the main challenge faced in thisarea. Moreover, its stability properties provide a natural connection between discrete dataand an underlying continuous structure. Finally, it can be combined with other tools, like thedistance to a measure, which allows to handle noise that are unbounded. The main caveat ofthis approach is its high complexity.In this thesis, we will introduce topological data analysis and persistent homology, then showhow to use approximation to reduce the computational complexity. We provide an approxima-tion scheme to the distance to a measure and a sparsifying method of weighted Vietoris-Ripscomplexes in order to approximate persistence diagrams with practical complexity. We detailthe specific properties of these constructions.Persistent homology was previously shown to be of use for scalar field analysis. We providea way to combine it with the distance to a measure in order to handle a wider class of noise,especially data with unbounded errors. Finally, we discuss interesting opportunities openedby these results to study data where parts are missing or erroneous.

M3 - Doctoral Thesis

ER -