Английская Википедия:Discrete Morse theory

Материал из Онлайн справочника
Перейти к навигацииПерейти к поиску

Discrete Morse theory is a combinatorial adaptation of Morse theory developed by Robin Forman. The theory has various practical applications in diverse fields of applied mathematics and computer science, such as configuration spaces,[1] homology computation,[2][3] denoising,[4] mesh compression,[5] and topological data analysis.[6]

Notation regarding CW complexes

Let <math>X</math> be a CW complex and denote by <math>\mathcal{X}</math> its set of cells. Define the incidence function <math>\kappa\colon\mathcal{X} \times \mathcal{X} \to \mathbb{Z}</math> in the following way: given two cells <math>\sigma</math> and <math>\tau</math> in <math>\mathcal{X}</math>, let <math>\kappa(\sigma,~\tau)</math> be the degree of the attaching map from the boundary of <math>\sigma</math> to <math>\tau</math>. The boundary operator is the endomorphism <math>\partial</math> of the free abelian group generated by <math>\mathcal{X}</math> defined by

<math>\partial(\sigma) = \sum_{\tau \in \mathcal{X}}\kappa(\sigma,\tau)\tau.</math>

It is a defining property of boundary operators that <math>\partial\circ\partial \equiv 0</math>. In more axiomatic definitions[7] one can find the requirement that <math>\forall \sigma,\tau^{\prime} \in \mathcal{X}</math>

<math> \sum_{\tau \in \mathcal{X}} \kappa(\sigma,\tau) \kappa(\tau,\tau^{\prime}) = 0</math>

which is a consequence of the above definition of the boundary operator and the requirement that <math>\partial\circ\partial \equiv 0</math>.

Discrete Morse functions

A real-valued function <math>\mu\colon\mathcal{X} \to \mathbb{R}</math> is a discrete Morse function if it satisfies the following two properties:

  1. For any cell <math>\sigma \in \mathcal{X}</math>, the number of cells <math>\tau \in \mathcal{X}</math> in the boundary of <math>\sigma</math> which satisfy <math>\mu(\sigma) \leq \mu(\tau)</math> is at most one.
  2. For any cell <math>\sigma \in \mathcal{X}</math>, the number of cells <math>\tau \in \mathcal{X}</math> containing <math>\sigma</math> in their boundary which satisfy <math>\mu(\sigma) \geq \mu(\tau)</math> is at most one.

It can be shown[8] that the cardinalities in the two conditions cannot both be one simultaneously for a fixed cell <math>\sigma</math>, provided that <math>\mathcal{X}</math> is a regular CW complex. In this case, each cell <math>\sigma \in \mathcal{X}</math> can be paired with at most one exceptional cell <math>\tau \in \mathcal{X}</math>: either a boundary cell with larger <math>\mu</math> value, or a co-boundary cell with smaller <math>\mu</math> value. The cells which have no pairs, i.e., whose function values are strictly higher than their boundary cells and strictly lower than their co-boundary cells are called critical cells. Thus, a discrete Morse function partitions the CW complex into three distinct cell collections: <math>\mathcal{X} = \mathcal{A} \sqcup \mathcal{K} \sqcup \mathcal{Q}</math>, where:

  1. <math>\mathcal{A}</math> denotes the critical cells which are unpaired,
  2. <math>\mathcal{K}</math> denotes cells which are paired with boundary cells, and
  3. <math>\mathcal{Q}</math> denotes cells which are paired with co-boundary cells.

By construction, there is a bijection of sets between <math>k</math>-dimensional cells in <math>\mathcal{K}</math> and the <math>(k-1)</math>-dimensional cells in <math>\mathcal{Q}</math>, which can be denoted by <math>p^k\colon\mathcal{K}^k \to \mathcal{Q}^{k-1}</math> for each natural number <math>k</math>. It is an additional technical requirement that for each <math>K \in \mathcal{K}^k</math>, the degree of the attaching map from the boundary of <math>K</math> to its paired cell <math>p^k(K) \in \mathcal{Q}</math> is a unit in the underlying ring of <math>\mathcal{X}</math>. For instance, over the integers <math>\mathbb{Z}</math>, the only allowed values are <math>\pm 1</math>. This technical requirement is guaranteed, for instance, when one assumes that <math>\mathcal{X}</math> is a regular CW complex over <math>\mathbb{Z}</math>.

The fundamental result of discrete Morse theory establishes that the CW complex <math>\mathcal{X}</math> is isomorphic on the level of homology to a new complex <math>\mathcal{A}</math> consisting of only the critical cells. The paired cells in <math>\mathcal{K}</math> and <math>\mathcal{Q}</math> describe gradient paths between adjacent critical cells which can be used to obtain the boundary operator on <math>\mathcal{A}</math>. Some details of this construction are provided in the next section.

The Morse complex

A gradient path is a sequence of paired cells

<math>\rho = (Q_1, K_1, Q_2, K_2, \ldots, Q_M, K_M)</math>

satisfying <math>Q_m = p(K_m)</math> and <math>\kappa(K_m,~Q_{m+1}) \neq 0</math>. The index of this gradient path is defined to be the integer

<math>\nu(\rho) = \frac{\prod_{m=1}^{M-1}-\kappa(K_m,Q_{m+1})}{\prod_{m=1}^{M}\kappa(K_m,Q_m)}.</math>

The division here makes sense because the incidence between paired cells must be <math>\pm 1</math>. Note that by construction, the values of the discrete Morse function <math>\mu</math> must decrease across <math>\rho</math>. The path <math>\rho</math> is said to connect two critical cells <math>A,A' \in \mathcal{A}</math> if <math>\kappa(A,Q_1) \neq 0 \neq \kappa(K_M,A')</math>. This relationship may be expressed as <math>A \stackrel{\rho}{\to} A'</math>. The multiplicity of this connection is defined to be the integer <math>m(\rho) = \kappa(A,Q_1)\cdot\nu(\rho)\cdot\kappa(K_M,A')</math>. Finally, the Morse boundary operator on the critical cells <math>\mathcal{A}</math> is defined by

<math>\Delta(A) = \kappa(A,A') + \sum_{A \stackrel{\rho}{\to} A'}m(\rho) A'</math>

where the sum is taken over all gradient path connections from <math>A</math> to <math>A'</math>.

Basic Results

Many of the familiar results from continuous Morse theory apply in the discrete setting.

The Morse Inequalities

Let <math>\mathcal{A}</math> be a Morse complex associated to the CW complex <math>\mathcal{X}</math>. The number <math>m_q = |\mathcal{A}_q|</math> of <math>q</math>-cells in <math>\mathcal{A}</math> is called the <math>q</math>-th Morse number. Let <math>\beta_q</math> denote the <math>q</math>-th Betti number of <math>\mathcal{X}</math>. Then, for any <math>N > 0</math>, the following inequalities[9] hold

<math>m_N \geq \beta_N</math>, and
<math>m_N - m_{N-1} + \dots \pm m_0 \geq \beta_N - \beta_{N-1} + \dots \pm \beta_0</math>

Moreover, the Euler characteristic <math>\chi(\mathcal{X})</math> of <math>\mathcal{X}</math> satisfies

<math>\chi(\mathcal{X}) = m_0 - m_1 + \dots \pm m_{\dim \mathcal{X}}</math>

Discrete Morse Homology and Homotopy Type

Let <math>\mathcal{X}</math> be a regular CW complex with boundary operator <math>\partial</math> and a discrete Morse function <math>\mu\colon\mathcal{X} \to \mathbb{R}</math>. Let <math>\mathcal{A}</math> be the associated Morse complex with Morse boundary operator <math>\Delta</math>. Then, there is an isomorphism[10] of homology groups

<math>H_*(\mathcal{X},\partial) \simeq H_*(\mathcal{A},\Delta),</math>

and similarly for the homotopy groups.

Applications

Discrete Morse theory finds its application in molecular shape analysis,[11] skeletonization of digital images/volumes,[12] graph reconstruction from noisy data,[13] denoising noisy point clouds[14] and analysing lithic tools in archaeology.[15]

See also

References

Шаблон:Reflist Шаблон:Refbegin

Шаблон:Refend