Persistence Minimal Free Lie Model
Siheng Yi · 2025
Read Persistence Minimal Free Lie Model by Siheng Yi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-08373.txt.
Introduction
Siheng Yi fuses two algebraic toolkits into one model for spaces that change over time. Quillen's minimal model describes a rational space with a free Lie algebra, a compact algebraic shadow of its homotopy. Persistence modules, meanwhile, track how homology evolves across a filtration, the core idea of topological data analysis. Yi builds a persistence version of the minimal Quillen model from these parts.
Stability results for the new model show it rests on firm ground. The framework promises a finer read on filtered data than homology groups alone can give. It also ties rational homotopy theory to data analysis in a reusable way.
Catalog record
| CREATOR | Siheng Yi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-08373
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Persistence Minimal Free Lie Model Thanks: ∗ Siheng Yi (E-mail: [email protected] ) Thanks: Department of Mathematics, Southern University of Science and Technology Siheng Yi Abstract . The minimal Quillen model is a free Lie model for rational spaces proposed by Quillen. Meanwhile, persistence modules are theoretical abstractions of persistent homology. In this paper, we integrate the ideas of rational homotopy theory and persistence modules to construct the persistence minimal Quillen model and discuss its stability. Our results provide a new algebraic framework for topological data analysis, which is more refined compared to directly computing the homology groups of the filtration of simplicial complexes. Furthermore, the stability results for persistence minimal Lie models ensure that our model is well-founded. 2020 Mathematics Subject Classification 55N31, 55P62 1. Introduction 1.1. background Persistence modules [ 17 , 6 , 15 ] , as algebraic structures encoding the evolution of topological features across scales, have become central to the mathematical framework of topological data analysis(TDA) [ 8 , 17 , 5 , 7 ] . Their development, applications, and theoretical richness bridge pure mathematics, computational topology, and data science. Rational homotopy theory is a branch of algebraic topology that studies the homotopy-theoretic properties of spaces by rationalizing their homotopy groups and focusing on computational methods. It simplifies ordinary homotopy theory by discarding torsion information and retaining only rational coefficients, enabling explicit calculations for spaces like simply connected finite CW complexes. A key innovation in this field is Daniel Quillen’s introduction of free Lie models, constructed through differential graded free Lie algebras, which encode the rational homotopy type of a space via its differential forms and Lie bracket. In persistent homology, if we have already determined the method for constructing simplicial complexes from point clouds, then the remaining issue is to establish algebraic models for these simplicial complexes. Currently, the most frequently used algebraic model is the homology groups over some coefficient field \Bbbk for simplicial complexes. When we specify the coefficient field to be a field of characteristic 0, we can employ rational homotopy theory to establish a more refined algebraic model for simplicial complexes. In this paper, we first introduce our main results in Section 1. In Section 2, we will review some basic knowledge about the rational homotopy theory and persistence modules. Finally, we will fully elaborate on our results and proof in Section 3. 1.2. Main Results Before presenting our results, we need to establish some notation. • DGL is the category of connected differential graded Lie algebras over \mathbb{Q} , • Ho ( DGL ) is the homotopy category of DGL , • Vec is the category of finite-dimensional vector spaces over \mathbb{Q} , • grVec is the category of graded finite-dimensional vector spaces over \mathbb{Q} , • Top is the category of simply-connected rational spaces of finite type, • d_{HI} is the homotopy interleaving distance [ 1 ] . In this paper, we define persistence minimal Quillen models, which are persistence minimal free Lie models for rational \mathbb{R} -spaces, where Top is the category of simply connected rational spaces of finite type and a rational \mathbb{R} -space is a functor (\mathbb{R},\leq)\to\textbf{Top} . Meanwhile, we prove the existence of persistence minimal Quillen models, that is Theorem 1.1 . For any rational \mathbb{R} -space \mathbb{X}:(\mathbb{R},\leq)\to\textbf{Top} , there exists a persistence minimal Quillen model M_{Qui}(\mathbb{X}):(\mathbb{R},\leq)\to\textbf{Ho}(\textbf{DGL}) such that M_{Qui}(\mathbb{X})_{t} is a minimal Quillen model of \mathbb{X}_{r} and M_{Qui}(\mathbb{X})(s\leq t) is a Lie representative of \mathbb{X}(s\leq t) up to weak equivalences. Persistence minimal Quillen models As persistence minimal Quillen models, which is a type of persistence modules, we can consider its stability results, and thus we have proven the following theorem. Theorem 1.2 . For any rational \mathbb{R} -spaces \mathbb{X} and \mathbb{Y} , we have • d_{I}^{\textbf{Ho}(\textbf{DGL})}(M_{Qui}(\mathbb{X}),M_{Qui}(\mathbb{Y}))\leq d_{HI}(\mathbb{X},\mathbb{Y})\leq d_{I}(\mathbb{X},\mathbb{Y}) • \begin{array}[]{ll}d_{I}^{\textbf{grVec}}(\pi_{*}(\mathbb{X}),\pi_{*}(\mathbb{Y}))&=d_{I}^{\textbf{grVec}}(H_{*}\circ M_{Qui}(\mathbb{X}),H_{*}\circ M_{Qui}(\mathbb{Y}))\\ &\leq d_{I}^{\textbf{Ho}(\textbf{DGL})}(M_{Qui}(\mathbb{X}),M_{Qui}(\mathbb{Y}))\end{array} • d_{I}^{\textbf{grVec}}(H_{*}(\mathbb{X}),H_{*}(\mathbb{Y}))=d_{I}^{\textbf{grVec}}(\mathbb{V},\mathbb{W})\leq d_{I}^{\textbf{Ho}(\textbf{DGL})}(M_{Qui}(\mathbb{X}),M_{Qui}(\mathbb{Y})) 2. Preliminaries 2.1. Interleaving Distance Persistence modules are the categorization [ 2 ] of persistent homology. In general, a persistence module can be defined as a functor F:\mathcal{P}\to C , in which \mathcal{P} is a poset, where the \text{ob }\mathcal{P} is the set \mathcal{P} , and morphisms are the partial order of \mathcal{P} , and C is any abelian category. It is obvious that the category \mathcal{P} is thin. A category is said to be thin if, for every pair of objects in this category, there exists at most one morphism from a to b . In specific studies, the poset \mathcal{P} that we usually consider is (\mathbb{R},\leq) or (\mathbb{Z},\leq) . In category \mathbb{R} , the object is a real number r\in\mathbb{R} , and the morphism r\to s exists if and only if r\leq s . Similarly, we can define the category (\mathbb{Z},\leq) . In this paper, persistence modules we discuss are functors (\mathbb{R},\leq)\to C , where C can be the category Vec , the category DGL , or the category Top . In this subsection, we will introduce morphisms and ’distance’ between persistence modules, that is, the interleaving distance d_{I} . The interleaving distance [ 9 ] between persistence modules can be seen as a generalization of the bottleneck distance d_{B} between persistence diagrams [ 8 ] . Note that unless otherwise specified, the persistence modules considered are always functors \mathbb{R}\to C , in which C is any abelian category. For a persistence module \mathbb{X}:\mathcal{P}\to C , we denote \mathbb{X}(a\leq b) as \mathbb{X}_{a\leq b} and denote \mathbb{X}(a) as \mathbb{X}_{a} . Definition 2.1 . For persistence modules \mathbb{X} and \mathbb{Y} , a morphism between \mathbb{X} and \mathbb{Y} is a natural transformation between \mathbb{X} and \mathbb{Y} , f:\mathbb{X}\Rightarrow\mathbb{Y} that is denoted as f:\mathbb{X}\to\mathbb{Y} . The collection of all functors from \mathcal{P} to C and all natural transformations between the functors is the category C^{\mathcal{P}} . If \mathcal{P}=\mathbb{R} , we may think that persistence modules depict the evolution of objects in C over time. For instance, if persistence modules \mathbb{X},\mathbb{Y} satisfying \mathbb{X}(t)=\mathbb{Y}(t+\delta) for some constant \delta , then \mathbb{X} and \mathbb{Y} are same by shifting time \delta . However, there is no isomorphism between persistence modules \mathbb{X} and \mathbb{Y} , even morphism generally. Then, we need to expand the notations of morphisms and isomorphisms between persistence modules to the new version that contains the information of \epsilon -shifting. For \delta\geq 0 , we define that the \delta -interleaving category I^{\delta} is the thin category such that \text{ob }I^{\delta}:=\mathbb{R}\times\{0,1\} and there is the morphism (r,i)\to(s,j) if and only if either (1) r+\delta\leq s , or (2) i=j and r\leq s . There exist two functors E^{0},E^{1}:\mathbb{R}\to I^{\delta} mapping r\in\mathbb{R} to (r,0) and (r,1) , respectively. Definition 2.2 . Let C be any category and \mathbb{X},\mathbb{Y}:\mathbb{R}\to C be any two functors. A \delta -interleaving between \mathbb{X} and \mathbb{Y} is a functor Z:I^{\delta}\to C satisfying Z\circ E^{0}=\mathbb{X} and Z\circ E^{1}=\mathbb{Y} . We call persistence modules \mathbb{X},\mathbb{Y}:\mathbb{R}\to C are \delta -interleaved, if there exists a functor Z:I^{\delta}\to C that is a \delta -interleaving between \mathbb{X} and \mathbb{Y} . Let \mathbb{X}(\delta):\mathbb{R}\to C be the functor by shifting \mathbb{X} downward by \delta , i.e., \mathbb{X}(\delta)_{r}:=\mathbb{X}_{r+\delta} and \mathbb{X}(\delta)_{r\leq s}:=\mathbb{X}_{r+\delta\leq s+\delta} for all r\leq s\in\mathbb{R} . Similarly, f(\delta):\mathbb{X}(\delta)\to\mathbb{Y}(\delta) is defined by f(\delta)_{t}:=f_{t+\delta} , where f:\mathbb{X}\to\mathbb{Y} is a morphism between persistence modules. Specially, we define the morphism \phi^{\mathbb{X},\delta}:\mathbb{X}\to\mathbb{X}(\delta) for any \mathbb{X}:\mathbb{R}\to C , in which \phi^{\mathbb{X},\delta}_{t}=\mathbb{X}_{t\leq t+\delta} .