SlideShare a Scribd company logo
Topology for Computing
Chapter 4. Homology
Sangwoo Mo
TDA Study
October 16, 2017
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 1 / 37
Table of Contents
1 Motivation
2 Euler characteristic
3 Homotopy
4 Homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 2 / 37
Table of Contents
1 Motivation
2 Euler characteristic
3 Homotopy
4 Homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 3 / 37
Classifying Topological Spaces
How to classify topological spaces?
1-manifold has only two types: line and loop
How about higher-dimension manifolds?
The natural way is equivalence class of homeomorphism
Homeomorphism preserves topology by definition
However, it is hard to find homeomorphism in general
Thus, we will classify spaces by topological invariants e.g.
Euler characteristic
Homotopy
Homology (our main focus)
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 4 / 37
Table of Contents
1 Motivation
2 Euler characteristic
3 Homotopy
4 Homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 5 / 37
Euler characteristic
Definition (Euler characteristic of simplicial complex)
Let K be a simplicial complex and si = card{σ ∈ K | dim σ = i}. The
Euler characteristic of K is
χ(K) =
dim K
i=0
(−1)i
si =
σ∈K−{∅}
(−1)dim σ
.
Definition (Euler characteristic of manifold)
Given any triangulation K of a manifold M, χ(K) is invariant. We call
χ(M) = χ(K) be the Euler characteristic of M.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 6 / 37
Euler characteristic
Example (basic 2-manifolds)
Here are some popular 2-manifolds. Sphere S2, torus T2, real projective
plane RP2
, and the Klein bottle K2, respectively.
FYI 1: M¨obius strip is homeomorphic to one point removed RP2
.
FYI 2: Klein bottle is homeomorphic to connected sum of two RP2
s.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 7 / 37
Euler characteristic
Example (χ of basic 2-manifolds)
We can compute the Euler characteristic of basic 2-manifolds by
triangulation of the diagram.
Moreover, we can extend this result to the general closed 2-manifolds by
connected sum of basic 2-manifolds.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 8 / 37
Connected Sum
Definition (connected sum)
The connected sum of two n-manifolds M1, M2 is
M1 # M2 = M1 − ˚Dn
1
∂˚Dn
1 =∂˚Dn
2
M2 − ˚Dn
2
where ˚Dn
1 , ˚Dn
2 are n-dimensional closed disks in M1, M2, respectively.
Example (connected sum of torus)
Connected sum of 2 tori. We call it genus 2 torus.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 9 / 37
Connected Sum
Theorem (χ of connected sum)
For compact surfaces M1, M2,
χ(M1 # M2) = χ(M1) + χ(M2) − 2.
Corollary (χ of connected sum of T2
/RP2
)
As corollary, χ(gT2) = 2 − 2g and χ(gRP2
) = 2 − g.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 10 / 37
Classification of 2-manifolds
We can prove that any closed surface are homeomorphic to the connected
sum of T2 or the connected sum of RP2
. (Sphere is genus 0 torus.)
Theorem (classification of 2-manifolds)
The closed 2-manifold M is homeomorphic to
the connected sum of T2 if it is orientable
the connected sum of RP2
if it is not orientable
Using Euler characteristic, we can also identify the number of connection.
Homeomorphism (and Euler characteristic) suceed to classify 2-manifolds.
However, we need better tools to classify higher-dimension manifolds.
Before we move on, we will introduce basic category theory to formally
define invariant, from topological space to algebraic structure.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 11 / 37
Basic Category Theory
Definition (category)
A category C consists of:
(a) a collection Ob(C) of objects
(b) sets Mor(X, Y ) of morphisms for each pair X, Y ∈ Ob(C); including a
distinguished identity morphism 1 = 1X ∈ Mor(X, Y ) for each X
(c) a composition of morphisms function ◦ : Mor(X, Y ) × Mor(Y , Z)
→ Mor(X, Z) for each triple X, Y , Z ∈ Ob(C), satisfying
f ◦ 1 = 1 ◦ f = f , and (f ◦ g) ◦ h = f ◦ (g ◦ h)
Definition (functor)
A (covariant) functor F from a category C to a category D assigns
(a) object X ∈ C to an object F(X) ∈ D
(b) morphism f ∈ Mor(X, Y ) to a morphism F(f ) ∈ Mor(F(X), F(Y ));
such that F(1) = 1 and F(f ◦ g) = F(f ) ◦ F(g)
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 12 / 37
Basic Category Theory
Example (examples of category)
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 13 / 37
Basic Category Theory
Example (visual illustration of functor)
F is a functor mapping category of 2-manifolds to category of 1-manifolds.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 14 / 37
Functor from Topology to Algebra
Transform difficult topological object to well-known algebraic object.
object: topological space → group & module
morphism: homeomorphism → isomorphism (not really!)
Homotopy and homology are easier to compute, but less discriminative
homology ⊂ homotopy ⊂ homeomorphy
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 15 / 37
Table of Contents
1 Motivation
2 Euler characteristic
3 Homotopy
4 Homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 16 / 37
Homotopy
Definition (homotopy)
The homotopy of two continuous functions f , g : X → Y is a continuous
function H : X × [0, 1] → Y such that H(x, 0) = f (x) and H(x, 1) = g(x)
for x ∈ X. We can think H as a continuous deformation of f into g.
Example (homotopy vs homeomorphy)
X and Y are homotopy equivalent, but not homeomorphic.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 17 / 37
Homotopy
Definition (homotopy group)
Choose base point s ∈ Sn and x ∈ X. The homotopy group πn(X, x) is
the set of homotopy classes of maps f : (Sn, s) → (X, x). For simply
connected space, πn(X, x) is independent to x, thus we can write πn(X).
We call π1(X) be the fundamental group.
Remark (homotopy ‘group’)
Remark that πn(X) has a group structure. Even more, for n ≥ 2, πn(X)
becomes an abelian group.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 18 / 37
Homotopy
Intuitively, homotopy group counts the number of loops
Example (fundamental group of torus)
The fundamental group of torus is π1(T2) = Z × Z.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 19 / 37
Undecidability of Homeomorphy/Homotopy
Theorem (undecidability of homeomorphy/homotopy)
For manifolds with dimension higher than 4, homeomorphy/homotopy
problem is undecidable.
Proof sketch (Markov’s construction).
For given group, construct a corresponding manifold using Dehn surgery.
It reduces the homeomorphy/homotopy problem to the group isomorphism
problem, which is known to be undecidable.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 20 / 37
Limitation of Homotopy
Homotopy group is one way to classify topological spaces
However, it has several problems:
not combinatorial
hard to compute
infinite description of πn(X)
⇒ It motivates the homology, the combinatorial computable functor with
finite description
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 21 / 37
Table of Contents
1 Motivation
2 Euler characteristic
3 Homotopy
4 Homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 22 / 37
Homology
There are several ways to define homology (e.g. CW-complex), but we will
only cover simplicial homology; We only consider a simplicial complex K
First, we will define chain and cycle, the extension of path and loop
Definition (chain group)
The kth chain group Ck(K), + of K is the free abelian group on the
oriented k-simplices. An element of Ck(K) is a k-chain,
c =
q
nq[σq], nq ∈ Z, σq ∈ K.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 23 / 37
Homology
Definition (boundary homomorphism)
The boundary homomorphism ∂k : Ck(K) → Ck−1(K) is
∂kσ =
i
(−1)i
[v0, v1, ..., ˆvi , ..., vn]
where ˆvi indicates that vi is deleted from the sequence.
Example (boundary homomorphism of simplices)
∂1[a, b] = b − a
∂2[a, b, c] = [b, c] − [a, c] + [a, b]
∂3[a, b, c, d] = [b, c, d] − [a, c, d] + [a, b, d] − [a, b, c]
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 24 / 37
Homology
Definition (chain complex)
The chain complex C∗ is a sequence of chain groups
0 −→ Cn
∂n
−→ Cn−1 −→ ... −→ C1
∂1
−→ C0
∂0
−→ 0
Definition (cycle, boundary)
The kth cycle group is Zk = ker ∂k and the kth boundary group is
Bk = im ∂k+1. Elements of Zk and Bk are cycle and boundary.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 25 / 37
Homology
Theorem (nesting of cycle/boundary group)
Zk and Bk are free abelian normal subgroup of Ck such that
Bk ≤ Zk ≤ Ck. Here, Bk ≤ Zk since ∂k−1∂k = 0.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 26 / 37
Homology
Definition (homology group)
The kth homology group is
Hk = Zk/Bk = ker ∂k/im ∂k−1.
Note that homology groups are finitely generated abelian.
Definition (Betti number)
The kth Betti number βk = βk(Hk) is the rank of the free part of Hk.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 27 / 37
Understanding Homology
For simplicity, we only consider subcomplexes of S3
Subcomplexes of S3 easy to handle (torsion-free)
Algorithmically, simple one point compactification converts R3 to S3
For 3-dim torsion-free space, Betti numbers have intuitive meaning
Alexander Duality:
β0 measures the number of components
β1 measures the number of tunnels
β0 measures the number of voids
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 28 / 37
Understanding Homology
Example (Alexander duality)
Homology of basic 2-manifolds.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 29 / 37
Invariance
We hope homology be invariant for same underlying space
First approach was proving that any two triangulations of a
topological space have a common refinement (Hauptvermutung)
However, it was false for manifolds of dimension ≥ 5
For general spaces, singular homology is introduced
Simplical homology is equivalent to singular homology
Algorithmically, we can just compute (easy) simplical homology
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 30 / 37
Relation to Euler characteristic
The sequence of homology groups also forms a chain complex
0 −→ Hn −→ Hn−1 −→ ... −→ H1 −→ H0 −→ 0
Definition (Euler charactersitic of chain complex)
The Euler charactersitic of chain complex C∗ is
χ(C∗) =
i
(−1)i
rank(Ci ).
Theorem (Euler-Poincar´e formula)
χ(C∗) = χ(H∗(C∗)) =
i
(−1)i
βi .
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 31 / 37
Relation to Homotopy
Theorem (Hurewicz theorem)
For any space X and positive integer k there is a group homomorphism
h∗ : πk(X) → Hk(X).
If X is (n − 1)-connected, h∗ is an isomorphism for all 2 ≤ k ≤ n and
abelinization for n = 1. In particular, the theorem says that
H1(X) ∼= π1(X)/[π1(X), π1(X)].
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 32 / 37
Homology with coefficients
We assumed chain group be the free abelian group, which is Z-module
We may replace this ring with any PID D, such as Z2
Definition (homology with coefficients)
The kth homology with ring of coefficients D is
Hk(K; D) = Zk(K; D)/Bk(K; D).
Our goal is to compute Hk(K; D) for arbitrary D, using Hk(K)
If we use Z2, the computation becomes greatly simple (see Chap 7.)
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 33 / 37
Homology with coefficients
Definition (short exact sequence)
The short exact sequence is seqeunce of groups
0
ϕ3
−→ A
ϕ2
−→ B
ϕ1
−→ C
ϕ0
−→ 0
such that im ϕi = ker ϕi−1.
Theorem (splitting lemma)
For short exact sequence, B ∼= A × C.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 34 / 37
Homology with coefficients
Theorem (universal coefficient theorem)
Let G be an Abelian. Then following sequence is short exact
0 → Hk(K) ⊗ G → Hk(K; G) → Hk−1(K) ∗ G → 0
where ⊗ is tensor product and ∗ is torsion product.
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 35 / 37
Homology with coefficients
Using universal coefficient theorem and splitting lemma, we can derive
Hk(K; D) for some special cases
Example (homology with coefficients)
1. Case Hk(K; Zp):
Hk(K; Zp) ∼= (Hk(K; Zp) ⊗ Zp) × (Hk(K; Zp) ∗ Zp)
∼= torsion part × (Zp)βk
2. Case Hk(K; F):
Hk(K; F) ∼= (Hk(K; F) ⊗ F) × (Hk(K; F) ∗ F) ∼= Fβk
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 36 / 37
Questions?
Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 37 / 37

More Related Content

PDF
Tutorial of topological_data_analysis_part_1(basic)
PDF
013_20160328_Topological_Measurement_Of_Protein_Compressibility
PDF
project
PPTX
Section 11: Normal Subgroups
PPT
DM-unit-3.ppt
PPTX
Topology M.Sc. 2 semester Mathematics compactness, unit - 4
PPTX
topology definitions
PPTX
TOPOLOGY and TYPES OF TOPOLOGY PowerPoint
Tutorial of topological_data_analysis_part_1(basic)
013_20160328_Topological_Measurement_Of_Protein_Compressibility
project
Section 11: Normal Subgroups
DM-unit-3.ppt
Topology M.Sc. 2 semester Mathematics compactness, unit - 4
topology definitions
TOPOLOGY and TYPES OF TOPOLOGY PowerPoint

What's hot (20)

PPTX
PRESENTATION ON INTRODUCTION TO SEVERAL VARIABLES AND PARTIAL DERIVATIVES
PPTX
Unit 1: Topological spaces (its definition and definition of open sets)
PDF
Fractional calculus and applications
PDF
Introduction of Partial Differential Equations
PDF
統計的学習の基礎 5章前半(~5.6)
PDF
Wasserstein GAN 수학 이해하기 I
PPT
Application of integral calculus
PDF
2.2 Properties of union, intersection and complement
PDF
Transformation of random variables
PPTX
GROUP AND SUBGROUP PPT 20By SONU KUMAR.pptx
PPTX
5.1 anti derivatives
PPT
10 parametric eequations of lines
PDF
カーネル法:正定値カーネルの理論
PPTX
십분딥러닝_16_WGAN (Wasserstein GANs)
PPT
27 triple integrals in spherical and cylindrical coordinates
PPT
Lesson 14 a - parametric equations
PDF
PRML エビデンス近似 3.5 3.6.1
PPT
Fuzzylogic
PPTX
Real analysis
PDF
PRML 2.3節 - ガウス分布
PRESENTATION ON INTRODUCTION TO SEVERAL VARIABLES AND PARTIAL DERIVATIVES
Unit 1: Topological spaces (its definition and definition of open sets)
Fractional calculus and applications
Introduction of Partial Differential Equations
統計的学習の基礎 5章前半(~5.6)
Wasserstein GAN 수학 이해하기 I
Application of integral calculus
2.2 Properties of union, intersection and complement
Transformation of random variables
GROUP AND SUBGROUP PPT 20By SONU KUMAR.pptx
5.1 anti derivatives
10 parametric eequations of lines
カーネル法:正定値カーネルの理論
십분딥러닝_16_WGAN (Wasserstein GANs)
27 triple integrals in spherical and cylindrical coordinates
Lesson 14 a - parametric equations
PRML エビデンス近似 3.5 3.6.1
Fuzzylogic
Real analysis
PRML 2.3節 - ガウス分布
Ad

Similar to Topology for Computing: Homology (20)

PDF
HOMOLOGY AND HOMOTOPY THEORY PRESENTATION
PDF
Algebraic Topology - Hatcher.pdf
PDF
Senior_Thesis_Evan_Oman
PDF
Foundations Of Algebraic Topology 1, Second Printing Edition Samuel Eilenberg
PDF
Foundations Of Algebraic Topology Samuel Eilenberg Norman Steenrod
PDF
Foundations Of Algebraic Topology 1, Second Printing Edition Samuel Eilenberg
PDF
Cubical Homotopy Theory Brian A Munson Ismar Voli
PDF
IRJET- A Study about Fundamental Group in Algebraic Topology
PDF
Independence Complexes
PDF
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
PDF
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
PPT
Topology.ppt
PPT
Topology.ppt
PPT
Topology.ppt
PDF
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
PPT
Topology of DNA and Quantum Computing.ppt
PDF
Mod Two Homology and Cohomology (Jean Claude)
PDF
Homotopical Topology 2nd Edition Anatoly T Fomenko Dmitry B Fuchs
PDF
Topological Data Analysis and Persistent Homology
PDF
GROUPOIDS, LOCAL SYSTEMS AND DIFFERENTIAL EQUATIONS
HOMOLOGY AND HOMOTOPY THEORY PRESENTATION
Algebraic Topology - Hatcher.pdf
Senior_Thesis_Evan_Oman
Foundations Of Algebraic Topology 1, Second Printing Edition Samuel Eilenberg
Foundations Of Algebraic Topology Samuel Eilenberg Norman Steenrod
Foundations Of Algebraic Topology 1, Second Printing Edition Samuel Eilenberg
Cubical Homotopy Theory Brian A Munson Ismar Voli
IRJET- A Study about Fundamental Group in Algebraic Topology
Independence Complexes
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
Topology.ppt
Topology.ppt
Topology.ppt
Foundations of Algebraic Topology 1st Edition Samuel Eilenberg
Topology of DNA and Quantum Computing.ppt
Mod Two Homology and Cohomology (Jean Claude)
Homotopical Topology 2nd Edition Anatoly T Fomenko Dmitry B Fuchs
Topological Data Analysis and Persistent Homology
GROUPOIDS, LOCAL SYSTEMS AND DIFFERENTIAL EQUATIONS
Ad

More from Sangwoo Mo (20)

PDF
Brief History of Visual Representation Learning
PDF
Learning Visual Representations from Uncurated Data
PDF
Hyperbolic Deep Reinforcement Learning
PDF
A Unified Framework for Computer Vision Tasks: (Conditional) Generative Model...
PDF
Self-supervised Learning Lecture Note
PDF
Deep Learning Theory Seminar (Chap 3, part 2)
PDF
Deep Learning Theory Seminar (Chap 1-2, part 1)
PDF
Introduction to Diffusion Models
PDF
Object-Region Video Transformers
PDF
Deep Implicit Layers: Learning Structured Problems with Neural Networks
PDF
Learning Theory 101 ...and Towards Learning the Flat Minima
PDF
Sharpness-aware minimization (SAM)
PDF
Explicit Density Models
PDF
Score-Based Generative Modeling through Stochastic Differential Equations
PDF
Self-Attention with Linear Complexity
PDF
Meta-Learning with Implicit Gradients
PDF
Challenging Common Assumptions in the Unsupervised Learning of Disentangled R...
PDF
Generative Models for General Audiences
PDF
Bayesian Model-Agnostic Meta-Learning
PDF
Deep Learning for Natural Language Processing
Brief History of Visual Representation Learning
Learning Visual Representations from Uncurated Data
Hyperbolic Deep Reinforcement Learning
A Unified Framework for Computer Vision Tasks: (Conditional) Generative Model...
Self-supervised Learning Lecture Note
Deep Learning Theory Seminar (Chap 3, part 2)
Deep Learning Theory Seminar (Chap 1-2, part 1)
Introduction to Diffusion Models
Object-Region Video Transformers
Deep Implicit Layers: Learning Structured Problems with Neural Networks
Learning Theory 101 ...and Towards Learning the Flat Minima
Sharpness-aware minimization (SAM)
Explicit Density Models
Score-Based Generative Modeling through Stochastic Differential Equations
Self-Attention with Linear Complexity
Meta-Learning with Implicit Gradients
Challenging Common Assumptions in the Unsupervised Learning of Disentangled R...
Generative Models for General Audiences
Bayesian Model-Agnostic Meta-Learning
Deep Learning for Natural Language Processing

Recently uploaded (20)

PDF
Mobile App Security Testing_ A Comprehensive Guide.pdf
PDF
Spectral efficient network and resource selection model in 5G networks
PDF
Empathic Computing: Creating Shared Understanding
PDF
Advanced methodologies resolving dimensionality complications for autism neur...
PDF
Agricultural_Statistics_at_a_Glance_2022_0.pdf
PDF
Heart disease approach using modified random forest and particle swarm optimi...
PDF
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
PPT
Teaching material agriculture food technology
PDF
Diabetes mellitus diagnosis method based random forest with bat algorithm
PPTX
Programs and apps: productivity, graphics, security and other tools
PDF
Network Security Unit 5.pdf for BCA BBA.
PDF
Machine learning based COVID-19 study performance prediction
PDF
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
PDF
Assigned Numbers - 2025 - Bluetooth® Document
PDF
Approach and Philosophy of On baking technology
PDF
Getting Started with Data Integration: FME Form 101
PDF
Unlocking AI with Model Context Protocol (MCP)
PDF
Building Integrated photovoltaic BIPV_UPV.pdf
PPTX
OMC Textile Division Presentation 2021.pptx
Mobile App Security Testing_ A Comprehensive Guide.pdf
Spectral efficient network and resource selection model in 5G networks
Empathic Computing: Creating Shared Understanding
Advanced methodologies resolving dimensionality complications for autism neur...
Agricultural_Statistics_at_a_Glance_2022_0.pdf
Heart disease approach using modified random forest and particle swarm optimi...
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Teaching material agriculture food technology
Diabetes mellitus diagnosis method based random forest with bat algorithm
Programs and apps: productivity, graphics, security and other tools
Network Security Unit 5.pdf for BCA BBA.
Machine learning based COVID-19 study performance prediction
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
Assigned Numbers - 2025 - Bluetooth® Document
Approach and Philosophy of On baking technology
Getting Started with Data Integration: FME Form 101
Unlocking AI with Model Context Protocol (MCP)
Building Integrated photovoltaic BIPV_UPV.pdf
OMC Textile Division Presentation 2021.pptx

Topology for Computing: Homology

  • 1. Topology for Computing Chapter 4. Homology Sangwoo Mo TDA Study October 16, 2017 Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 1 / 37
  • 2. Table of Contents 1 Motivation 2 Euler characteristic 3 Homotopy 4 Homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 2 / 37
  • 3. Table of Contents 1 Motivation 2 Euler characteristic 3 Homotopy 4 Homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 3 / 37
  • 4. Classifying Topological Spaces How to classify topological spaces? 1-manifold has only two types: line and loop How about higher-dimension manifolds? The natural way is equivalence class of homeomorphism Homeomorphism preserves topology by definition However, it is hard to find homeomorphism in general Thus, we will classify spaces by topological invariants e.g. Euler characteristic Homotopy Homology (our main focus) Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 4 / 37
  • 5. Table of Contents 1 Motivation 2 Euler characteristic 3 Homotopy 4 Homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 5 / 37
  • 6. Euler characteristic Definition (Euler characteristic of simplicial complex) Let K be a simplicial complex and si = card{σ ∈ K | dim σ = i}. The Euler characteristic of K is χ(K) = dim K i=0 (−1)i si = σ∈K−{∅} (−1)dim σ . Definition (Euler characteristic of manifold) Given any triangulation K of a manifold M, χ(K) is invariant. We call χ(M) = χ(K) be the Euler characteristic of M. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 6 / 37
  • 7. Euler characteristic Example (basic 2-manifolds) Here are some popular 2-manifolds. Sphere S2, torus T2, real projective plane RP2 , and the Klein bottle K2, respectively. FYI 1: M¨obius strip is homeomorphic to one point removed RP2 . FYI 2: Klein bottle is homeomorphic to connected sum of two RP2 s. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 7 / 37
  • 8. Euler characteristic Example (χ of basic 2-manifolds) We can compute the Euler characteristic of basic 2-manifolds by triangulation of the diagram. Moreover, we can extend this result to the general closed 2-manifolds by connected sum of basic 2-manifolds. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 8 / 37
  • 9. Connected Sum Definition (connected sum) The connected sum of two n-manifolds M1, M2 is M1 # M2 = M1 − ˚Dn 1 ∂˚Dn 1 =∂˚Dn 2 M2 − ˚Dn 2 where ˚Dn 1 , ˚Dn 2 are n-dimensional closed disks in M1, M2, respectively. Example (connected sum of torus) Connected sum of 2 tori. We call it genus 2 torus. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 9 / 37
  • 10. Connected Sum Theorem (χ of connected sum) For compact surfaces M1, M2, χ(M1 # M2) = χ(M1) + χ(M2) − 2. Corollary (χ of connected sum of T2 /RP2 ) As corollary, χ(gT2) = 2 − 2g and χ(gRP2 ) = 2 − g. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 10 / 37
  • 11. Classification of 2-manifolds We can prove that any closed surface are homeomorphic to the connected sum of T2 or the connected sum of RP2 . (Sphere is genus 0 torus.) Theorem (classification of 2-manifolds) The closed 2-manifold M is homeomorphic to the connected sum of T2 if it is orientable the connected sum of RP2 if it is not orientable Using Euler characteristic, we can also identify the number of connection. Homeomorphism (and Euler characteristic) suceed to classify 2-manifolds. However, we need better tools to classify higher-dimension manifolds. Before we move on, we will introduce basic category theory to formally define invariant, from topological space to algebraic structure. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 11 / 37
  • 12. Basic Category Theory Definition (category) A category C consists of: (a) a collection Ob(C) of objects (b) sets Mor(X, Y ) of morphisms for each pair X, Y ∈ Ob(C); including a distinguished identity morphism 1 = 1X ∈ Mor(X, Y ) for each X (c) a composition of morphisms function ◦ : Mor(X, Y ) × Mor(Y , Z) → Mor(X, Z) for each triple X, Y , Z ∈ Ob(C), satisfying f ◦ 1 = 1 ◦ f = f , and (f ◦ g) ◦ h = f ◦ (g ◦ h) Definition (functor) A (covariant) functor F from a category C to a category D assigns (a) object X ∈ C to an object F(X) ∈ D (b) morphism f ∈ Mor(X, Y ) to a morphism F(f ) ∈ Mor(F(X), F(Y )); such that F(1) = 1 and F(f ◦ g) = F(f ) ◦ F(g) Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 12 / 37
  • 13. Basic Category Theory Example (examples of category) Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 13 / 37
  • 14. Basic Category Theory Example (visual illustration of functor) F is a functor mapping category of 2-manifolds to category of 1-manifolds. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 14 / 37
  • 15. Functor from Topology to Algebra Transform difficult topological object to well-known algebraic object. object: topological space → group & module morphism: homeomorphism → isomorphism (not really!) Homotopy and homology are easier to compute, but less discriminative homology ⊂ homotopy ⊂ homeomorphy Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 15 / 37
  • 16. Table of Contents 1 Motivation 2 Euler characteristic 3 Homotopy 4 Homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 16 / 37
  • 17. Homotopy Definition (homotopy) The homotopy of two continuous functions f , g : X → Y is a continuous function H : X × [0, 1] → Y such that H(x, 0) = f (x) and H(x, 1) = g(x) for x ∈ X. We can think H as a continuous deformation of f into g. Example (homotopy vs homeomorphy) X and Y are homotopy equivalent, but not homeomorphic. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 17 / 37
  • 18. Homotopy Definition (homotopy group) Choose base point s ∈ Sn and x ∈ X. The homotopy group πn(X, x) is the set of homotopy classes of maps f : (Sn, s) → (X, x). For simply connected space, πn(X, x) is independent to x, thus we can write πn(X). We call π1(X) be the fundamental group. Remark (homotopy ‘group’) Remark that πn(X) has a group structure. Even more, for n ≥ 2, πn(X) becomes an abelian group. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 18 / 37
  • 19. Homotopy Intuitively, homotopy group counts the number of loops Example (fundamental group of torus) The fundamental group of torus is π1(T2) = Z × Z. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 19 / 37
  • 20. Undecidability of Homeomorphy/Homotopy Theorem (undecidability of homeomorphy/homotopy) For manifolds with dimension higher than 4, homeomorphy/homotopy problem is undecidable. Proof sketch (Markov’s construction). For given group, construct a corresponding manifold using Dehn surgery. It reduces the homeomorphy/homotopy problem to the group isomorphism problem, which is known to be undecidable. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 20 / 37
  • 21. Limitation of Homotopy Homotopy group is one way to classify topological spaces However, it has several problems: not combinatorial hard to compute infinite description of πn(X) ⇒ It motivates the homology, the combinatorial computable functor with finite description Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 21 / 37
  • 22. Table of Contents 1 Motivation 2 Euler characteristic 3 Homotopy 4 Homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 22 / 37
  • 23. Homology There are several ways to define homology (e.g. CW-complex), but we will only cover simplicial homology; We only consider a simplicial complex K First, we will define chain and cycle, the extension of path and loop Definition (chain group) The kth chain group Ck(K), + of K is the free abelian group on the oriented k-simplices. An element of Ck(K) is a k-chain, c = q nq[σq], nq ∈ Z, σq ∈ K. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 23 / 37
  • 24. Homology Definition (boundary homomorphism) The boundary homomorphism ∂k : Ck(K) → Ck−1(K) is ∂kσ = i (−1)i [v0, v1, ..., ˆvi , ..., vn] where ˆvi indicates that vi is deleted from the sequence. Example (boundary homomorphism of simplices) ∂1[a, b] = b − a ∂2[a, b, c] = [b, c] − [a, c] + [a, b] ∂3[a, b, c, d] = [b, c, d] − [a, c, d] + [a, b, d] − [a, b, c] Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 24 / 37
  • 25. Homology Definition (chain complex) The chain complex C∗ is a sequence of chain groups 0 −→ Cn ∂n −→ Cn−1 −→ ... −→ C1 ∂1 −→ C0 ∂0 −→ 0 Definition (cycle, boundary) The kth cycle group is Zk = ker ∂k and the kth boundary group is Bk = im ∂k+1. Elements of Zk and Bk are cycle and boundary. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 25 / 37
  • 26. Homology Theorem (nesting of cycle/boundary group) Zk and Bk are free abelian normal subgroup of Ck such that Bk ≤ Zk ≤ Ck. Here, Bk ≤ Zk since ∂k−1∂k = 0. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 26 / 37
  • 27. Homology Definition (homology group) The kth homology group is Hk = Zk/Bk = ker ∂k/im ∂k−1. Note that homology groups are finitely generated abelian. Definition (Betti number) The kth Betti number βk = βk(Hk) is the rank of the free part of Hk. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 27 / 37
  • 28. Understanding Homology For simplicity, we only consider subcomplexes of S3 Subcomplexes of S3 easy to handle (torsion-free) Algorithmically, simple one point compactification converts R3 to S3 For 3-dim torsion-free space, Betti numbers have intuitive meaning Alexander Duality: β0 measures the number of components β1 measures the number of tunnels β0 measures the number of voids Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 28 / 37
  • 29. Understanding Homology Example (Alexander duality) Homology of basic 2-manifolds. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 29 / 37
  • 30. Invariance We hope homology be invariant for same underlying space First approach was proving that any two triangulations of a topological space have a common refinement (Hauptvermutung) However, it was false for manifolds of dimension ≥ 5 For general spaces, singular homology is introduced Simplical homology is equivalent to singular homology Algorithmically, we can just compute (easy) simplical homology Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 30 / 37
  • 31. Relation to Euler characteristic The sequence of homology groups also forms a chain complex 0 −→ Hn −→ Hn−1 −→ ... −→ H1 −→ H0 −→ 0 Definition (Euler charactersitic of chain complex) The Euler charactersitic of chain complex C∗ is χ(C∗) = i (−1)i rank(Ci ). Theorem (Euler-Poincar´e formula) χ(C∗) = χ(H∗(C∗)) = i (−1)i βi . Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 31 / 37
  • 32. Relation to Homotopy Theorem (Hurewicz theorem) For any space X and positive integer k there is a group homomorphism h∗ : πk(X) → Hk(X). If X is (n − 1)-connected, h∗ is an isomorphism for all 2 ≤ k ≤ n and abelinization for n = 1. In particular, the theorem says that H1(X) ∼= π1(X)/[π1(X), π1(X)]. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 32 / 37
  • 33. Homology with coefficients We assumed chain group be the free abelian group, which is Z-module We may replace this ring with any PID D, such as Z2 Definition (homology with coefficients) The kth homology with ring of coefficients D is Hk(K; D) = Zk(K; D)/Bk(K; D). Our goal is to compute Hk(K; D) for arbitrary D, using Hk(K) If we use Z2, the computation becomes greatly simple (see Chap 7.) Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 33 / 37
  • 34. Homology with coefficients Definition (short exact sequence) The short exact sequence is seqeunce of groups 0 ϕ3 −→ A ϕ2 −→ B ϕ1 −→ C ϕ0 −→ 0 such that im ϕi = ker ϕi−1. Theorem (splitting lemma) For short exact sequence, B ∼= A × C. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 34 / 37
  • 35. Homology with coefficients Theorem (universal coefficient theorem) Let G be an Abelian. Then following sequence is short exact 0 → Hk(K) ⊗ G → Hk(K; G) → Hk−1(K) ∗ G → 0 where ⊗ is tensor product and ∗ is torsion product. Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 35 / 37
  • 36. Homology with coefficients Using universal coefficient theorem and splitting lemma, we can derive Hk(K; D) for some special cases Example (homology with coefficients) 1. Case Hk(K; Zp): Hk(K; Zp) ∼= (Hk(K; Zp) ⊗ Zp) × (Hk(K; Zp) ∗ Zp) ∼= torsion part × (Zp)βk 2. Case Hk(K; F): Hk(K; F) ∼= (Hk(K; F) ⊗ F) × (Hk(K; F) ∗ F) ∼= Fβk Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 36 / 37
  • 37. Questions? Sangwoo Mo (TDA Study) Chap 4. Homology October 16, 2017 37 / 37