12.04.2014 Views

Introduction to Applied Algebraic Topology: Persistent Homology

Introduction to Applied Algebraic Topology: Persistent Homology

Introduction to Applied Algebraic Topology: Persistent Homology

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Introduction</strong> <strong>to</strong> <strong>Applied</strong><br />

<strong>Algebraic</strong> <strong>Topology</strong>:<br />

<strong>Persistent</strong> <strong>Homology</strong><br />

Joshua Roberts<br />

Department of Mathematics<br />

University of Kentucky<br />

September 23, 2009<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 1/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 2/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>? (The<br />

<strong>Topology</strong> Part)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 3/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>? (The<br />

<strong>Topology</strong> Part)<br />

• <strong>Topology</strong> - Study of shapes that can be deformed<br />

continuously in<strong>to</strong> other shapes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 3/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>? (The<br />

<strong>Topology</strong> Part)<br />

• <strong>Topology</strong> - Study of shapes that can be deformed<br />

continuously in<strong>to</strong> other shapes<br />

• The surface of a cube can be “flattened” <strong>to</strong> a sphere, they<br />

are homeomorphic (<strong>to</strong>pologically equivalent)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 3/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>? (The<br />

<strong>Topology</strong> Part)<br />

• <strong>Topology</strong> - Study of shapes that can be deformed<br />

continuously in<strong>to</strong> other shapes<br />

• The surface of a cube can be “flattened” <strong>to</strong> a sphere, they<br />

are homeomorphic (<strong>to</strong>pologically equivalent)<br />

...also...<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 3/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>? (The<br />

<strong>Topology</strong> Part)<br />

• <strong>Topology</strong> - Study of shapes that can be deformed<br />

continuously in<strong>to</strong> other shapes<br />

• The surface of a cube can be “flattened” <strong>to</strong> a sphere, they<br />

are homeomorphic (<strong>to</strong>pologically equivalent)<br />

...also...<br />

Click Me!!!<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 3/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

• Problem - Homeomorphism is <strong>to</strong>o strong!<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

• Problem - Homeomorphism is <strong>to</strong>o strong! Difficult <strong>to</strong><br />

classify <strong>to</strong>pological spaces up <strong>to</strong> homeomorphism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

• Problem - Homeomorphism is <strong>to</strong>o strong! Difficult <strong>to</strong><br />

classify <strong>to</strong>pological spaces up <strong>to</strong> homeomorphism<br />

• Solution - Homo<strong>to</strong>py!<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

• Problem - Homeomorphism is <strong>to</strong>o strong! Difficult <strong>to</strong><br />

classify <strong>to</strong>pological spaces up <strong>to</strong> homeomorphism<br />

• Solution - Homo<strong>to</strong>py! Intuitively, two spaces are homo<strong>to</strong>py<br />

equivalent if one can be deformed in<strong>to</strong> the other, allowing<br />

contractions and expansion<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Homo<strong>to</strong>py Equivalence<br />

• Problem - Homeomorphism is <strong>to</strong>o strong! Difficult <strong>to</strong><br />

classify <strong>to</strong>pological spaces up <strong>to</strong> homeomorphism<br />

• Solution - Homo<strong>to</strong>py! Intuitively, two spaces are homo<strong>to</strong>py<br />

equivalent if one can be deformed in<strong>to</strong> the other, allowing<br />

contractions and expansion<br />

• Ex:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 4/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

• <strong>Algebraic</strong> <strong>to</strong>pology assigns algebraic invariants <strong>to</strong><br />

<strong>to</strong>pological spaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

• <strong>Algebraic</strong> <strong>to</strong>pology assigns algebraic invariants <strong>to</strong><br />

<strong>to</strong>pological spaces.<br />

• χ - Euler characteristic<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

• <strong>Algebraic</strong> <strong>to</strong>pology assigns algebraic invariants <strong>to</strong><br />

<strong>to</strong>pological spaces.<br />

• χ - Euler characteristic - Assigns integers <strong>to</strong> <strong>to</strong>pological<br />

spaces<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

• <strong>Algebraic</strong> <strong>to</strong>pology assigns algebraic invariants <strong>to</strong><br />

<strong>to</strong>pological spaces.<br />

• χ - Euler characteristic - Assigns integers <strong>to</strong> <strong>to</strong>pological<br />

spaces - Homo<strong>to</strong>py Invariant<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

What is <strong>Algebraic</strong> <strong>Topology</strong>?<br />

• <strong>Algebraic</strong> <strong>to</strong>pology assigns algebraic invariants <strong>to</strong><br />

<strong>to</strong>pological spaces.<br />

• χ - Euler characteristic - Assigns integers <strong>to</strong> <strong>to</strong>pological<br />

spaces - Homo<strong>to</strong>py Invariant<br />

• If X is homo<strong>to</strong>py equivalent <strong>to</strong> a polyhedron,<br />

χ(X) = #v − #e + #f<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 5/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Euler characteristic<br />

Theorem<br />

The Euler characteristic completely classifies (compact,<br />

oriented) surfaces.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 6/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

“<strong>Topology</strong>! The stra<strong>to</strong>sphere of human thought! In the<br />

twenty-fourth century it might possibly be of use <strong>to</strong> someone...”<br />

The First Circle, A. Solzhenitsyn<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 7/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 8/ 42


Simplices<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 9/ 42


Simplices<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplex is the convex hull of a collection of points in<br />

“general position”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 9/ 42


Simplices<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplex is the convex hull of a collection of points in<br />

“general position”<br />

• A 4-simplex is ...<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 9/ 42


Simplices<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplex is the convex hull of a collection of points in<br />

“general position”<br />

• A 4-simplex is ... (you get the idea)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 9/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complexes<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplicial complex is a bunch of simplices glued <strong>to</strong>gether<br />

so that the intersection of any two simplicies is another<br />

simplex.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 10/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complexes<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplicial complex is a bunch of simplices glued <strong>to</strong>gether<br />

so that the intersection of any two simplicies is another<br />

simplex.<br />

Figure:<br />

Simplicial<br />

Complexes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 10/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complexes<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• A simplicial complex is a bunch of simplices glued <strong>to</strong>gether<br />

so that the intersection of any two simplicies is another<br />

simplex.<br />

Figure:<br />

Simplicial<br />

Complexes<br />

Figure: Not a<br />

Simplicial<br />

Complex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 10/ 42


Triangulations<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Theorem<br />

Every <strong>to</strong>pological space (manifold) up <strong>to</strong> dimension three is<br />

homeomorphic <strong>to</strong> a simplicial complex.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 11/ 42


Triangulations<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Theorem<br />

Every <strong>to</strong>pological space (manifold) up <strong>to</strong> dimension three is<br />

homeomorphic <strong>to</strong> a simplicial complex.<br />

This is called a triangulation of the space.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 11/ 42


Triangulations<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Theorem<br />

Every <strong>to</strong>pological space (manifold) up <strong>to</strong> dimension three is<br />

homeomorphic <strong>to</strong> a simplicial complex.<br />

This is called a triangulation of the space.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 11/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Let X be the simplicial complex ...<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Let X be the simplicial complex ...<br />

• C 0 (X) = F 2v × F 2w × F 2x × F 2y × F 2z<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Let X be the simplicial complex ...<br />

• C 0 (X) = F 2v × F 2w × F 2x × F 2y × F 2z<br />

• C 1 (X) = F 2a × F 2b × F 2c × F 2d × F 2e × F 2f<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Let X be the simplicial complex ...<br />

• C 0 (X) = F 2v × F 2w × F 2x × F 2y × F 2z<br />

• C 1 (X) = F 2a × F 2b × F 2c × F 2d × F 2e × F 2f<br />

• C 2 (X) = F 2K<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Complexes <strong>to</strong> Vec<strong>to</strong>r Spaces<br />

• Given a simplicial complex, we can associate vec<strong>to</strong>r<br />

spaces (over a chosen field) corresponding <strong>to</strong> n-simplicies,<br />

where a simplex is a basis element for the vec<strong>to</strong>r space.<br />

Let X be the simplicial complex ...<br />

• C 0 (X) = F 2v × F 2w × F 2x × F 2y × F 2z<br />

• C 1 (X) = F 2a × F 2b × F 2c × F 2d × F 2e × F 2f<br />

• C 2 (X) = F 2K<br />

• C n (X) = 0 for n ≥ 3<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 12/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 13/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

H n (X) = “Cycles”/“Boundaries”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

H n (X) = “Cycles”/“Boundaries”<br />

H 1 (X)<br />

= ([a + b + c] + [d + e + f ])/∂K<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

H n (X) = “Cycles”/“Boundaries”<br />

H 1 (X)<br />

= ([a + b + c] + [d + e + f ])/∂K<br />

= ([a + b + c] + [d + e + f ])/[a + b + c]<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

H n (X) = “Cycles”/“Boundaries”<br />

H 1 (X)<br />

= ([a + b + c] + [d + e + f ])/∂K<br />

= ([a + b + c] + [d + e + f ])/[a + b + c]<br />

∼= [d + e + f ]<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• Given a simplicial complex X, Define the<br />

n th homology of X <strong>to</strong> be the quotient<br />

vec<strong>to</strong>r space<br />

H n (X) = “Cycles”/“Boundaries”<br />

H 1 (X)<br />

= ([a + b + c] + [d + e + f ])/∂K<br />

= ([a + b + c] + [d + e + f ])/[a + b + c]<br />

∼= [d + e + f ]<br />

∼= F 2<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 14/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

• 0-dimensional holes are connected components<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

• 0-dimensional holes are connected components<br />

• 1-dimensional holes are circles<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

• 0-dimensional holes are connected components<br />

• 1-dimensional holes are circles<br />

• 2-dimensional holes are cavities<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

• 0-dimensional holes are connected components<br />

• 1-dimensional holes are circles<br />

• 2-dimensional holes are cavities<br />

• 3-dimensional holes are ...<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial <strong>Homology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

The intuition is that homology measures “holes”<br />

• 0-dimensional holes are connected components<br />

• 1-dimensional holes are circles<br />

• 2-dimensional holes are cavities<br />

• 3-dimensional holes are ... (you get the idea)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 15/ 42


Betti Numbers<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 16/ 42


Betti Numbers<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• The n th Betti number of a simplicial complex X is the<br />

dimension of the vec<strong>to</strong>r space H n (X).<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 16/ 42


Betti Numbers<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• The n th Betti number of a simplicial complex X is the<br />

dimension of the vec<strong>to</strong>r space H n (X).<br />

• This is also the number of n-dimensional holes in the<br />

simplex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 16/ 42


Betti Numbers<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• The n th Betti number of a simplicial complex X is the<br />

dimension of the vec<strong>to</strong>r space H n (X).<br />

• This is also the number of n-dimensional holes in the<br />

simplex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 16/ 42


Betti Numbers<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Simplicial Complex<br />

<strong>Homology</strong><br />

• The n th Betti number of a simplicial complex X is the<br />

dimension of the vec<strong>to</strong>r space H n (X).<br />

• This is also the number of n-dimensional holes in the<br />

simplex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 16/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 17/ 42


Point-Clouds<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 18/ 42


Point-Clouds<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Let X be a collection of points in Euclidean space R n<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 18/ 42


Point-Clouds<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Let X be a collection of points in Euclidean space R n<br />

• In practice, X is obtained from data coming from<br />

questionnaires, sensor readings, population size, etc<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 18/ 42


Point-Clouds<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Let X be a collection of points in Euclidean space R n<br />

• In practice, X is obtained from data coming from<br />

questionnaires, sensor readings, population size, etc<br />

• Assumption: The data come from a (potentially noisy)<br />

sampling of points from an underlying <strong>to</strong>pological space (or<br />

manifold)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 18/ 42


Point-Clouds<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Let X be a collection of points in Euclidean space R n<br />

• In practice, X is obtained from data coming from<br />

questionnaires, sensor readings, population size, etc<br />

• Assumption: The data come from a (potentially noisy)<br />

sampling of points from an underlying <strong>to</strong>pological space (or<br />

manifold)<br />

Figure: Planar projection of higher dimensional point cloud data set<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 18/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Topological Methods<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 19/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Topological Methods<br />

Point-Cloud Data<br />

Persistence<br />

• We analyze a point cloud data set by putting a simplicial<br />

structure on X and calculating H n (X)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 19/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Topological Methods<br />

Point-Cloud Data<br />

Persistence<br />

• We analyze a point cloud data set by putting a simplicial<br />

structure on X and calculating H n (X)<br />

• Čech complex & Rips complex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 19/ 42


Čech Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 20/ 42


Čech Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, put a ball of radius ɛ/2 around<br />

each point<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 20/ 42


Čech Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, put a ball of radius ɛ/2 around<br />

each point<br />

• The Čech complex, C ɛ , is the simplicial complex whose<br />

k-simplicies are k-fold intersections of the ɛ/2-balls<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 20/ 42


Čech Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, put a ball of radius ɛ/2 around<br />

each point<br />

• The Čech complex, C ɛ , is the simplicial complex whose<br />

k-simplicies are k-fold intersections of the ɛ/2-balls<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 20/ 42


Čech Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, put a ball of radius ɛ/2 around<br />

each point<br />

• The Čech complex, C ɛ , is the simplicial complex whose<br />

k-simplicies are k-fold intersections of the ɛ/2-balls<br />

Theorem (Nerve Lemma)<br />

The Čech complex has the homo<strong>to</strong>py type of the cover by<br />

ɛ/2-balls.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 20/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

• Each set of k points within ɛ of each other forms a<br />

(k − 1)-simplex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

• Each set of k points within ɛ of each other forms a<br />

(k − 1)-simplex<br />

• Two points within ɛ - edge<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

• Each set of k points within ɛ of each other forms a<br />

(k − 1)-simplex<br />

• Two points within ɛ - edge<br />

• Three points within ɛ - triangle<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

• Each set of k points within ɛ of each other forms a<br />

(k − 1)-simplex<br />

• Two points within ɛ - edge<br />

• Three points within ɛ - triangle<br />

• Four points - tetrahedron<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


Rips Complex<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given a set of points X, the Rips complex R ɛ is the<br />

simplicial complex formed by:<br />

• Each set of k points within ɛ of each other forms a<br />

(k − 1)-simplex<br />

• Two points within ɛ - edge<br />

• Three points within ɛ - triangle<br />

• Four points - tetrahedron<br />

• Etc...<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 21/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 22/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

C ɛ ≃ S 1 ∨ S 1 ∨ S 1 and R ɛ ≃ S 1 ∨ S 2<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 22/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

• Good:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

• Good: Combina<strong>to</strong>rially simpler ⇒ Computationally<br />

cheaper<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

• Good: Combina<strong>to</strong>rially simpler ⇒ Computationally<br />

cheaper<br />

• Bad:<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

• Good: Combina<strong>to</strong>rially simpler ⇒ Computationally<br />

cheaper<br />

• Bad: Not homo<strong>to</strong>py equivalent <strong>to</strong> union of balls<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


Čech vs. Rips<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Čech Complex<br />

• Good: Homo<strong>to</strong>py equivalent <strong>to</strong> the union of balls<br />

• Bad: Can grow very big ⇒ Computationally expensive<br />

Rips Complex<br />

• Good: Combina<strong>to</strong>rially simpler ⇒ Computationally<br />

cheaper<br />

• Bad: Not homo<strong>to</strong>py equivalent <strong>to</strong> union of balls<br />

To compensate we use (a version of) the Rips complex, but<br />

vary ɛ<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 23/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given point cloud data obtained from an annulus, we have<br />

a sequence of Rips complexes by increasing ɛ.<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 24/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• Given point cloud data obtained from an annulus, we have<br />

a sequence of Rips complexes by increasing ɛ.<br />

• The fundamental question is “Which holes are real<br />

<strong>to</strong>pological features?”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 24/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 25/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Persistent</strong> <strong>Homology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 26/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Persistent</strong> <strong>Homology</strong><br />

Point-Cloud Data<br />

Persistence<br />

• We formalize and attempt <strong>to</strong> answer with the concept of<br />

persistent homology<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 26/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Persistent</strong> <strong>Homology</strong><br />

Point-Cloud Data<br />

Persistence<br />

• We formalize and attempt <strong>to</strong> answer with the concept of<br />

persistent homology<br />

• Given a simplicial complex X, a filtration F of X is a<br />

sequence of subcomplexes such that<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 26/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

<strong>Persistent</strong> <strong>Homology</strong><br />

Point-Cloud Data<br />

Persistence<br />

• We formalize and attempt <strong>to</strong> answer with the concept of<br />

persistent homology<br />

• Given a simplicial complex X, a filtration F of X is a<br />

sequence of subcomplexes such that<br />

∅ = X 0 ⊂ X 1 ⊂ · · · ⊂ X n−1 ⊂ X n = X<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 26/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

For the simplicial complex<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 27/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

For the simplicial complex<br />

Point-Cloud Data<br />

Persistence<br />

A filtration is<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 27/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

For the simplicial complex<br />

Point-Cloud Data<br />

Persistence<br />

A filtration is<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 27/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

For the simplicial complex<br />

Point-Cloud Data<br />

Persistence<br />

A filtration is<br />

• We form the Rips complex at each stage 0 ≤ i ≤ n and<br />

look at the homology<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 27/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

For the simplicial complex<br />

Point-Cloud Data<br />

Persistence<br />

A filtration is<br />

• We form the Rips complex at each stage 0 ≤ i ≤ n and<br />

look at the homology<br />

• Topological features that persist are considered significant<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 27/ 42


Barcodes<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

We record the homology by barcodes displaying the Betti<br />

numbers for levels of our filtration<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 28/ 42


Barcodes<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

We record the homology by barcodes displaying the Betti<br />

numbers for levels of our filtration<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 28/ 42


Barcodes<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

We record the homology by barcodes displaying the Betti<br />

numbers for levels of our filtration<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 28/ 42


Barcodes<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

We record the homology by barcodes displaying the Betti<br />

numbers for levels of our filtration<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 28/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 29/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Suppose a data set X yields the following barcodes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 29/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Suppose a data set X yields the following barcodes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 29/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Suppose a data set X yields the following barcodes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 29/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

• Two persistent Betti numbers in dimension 1<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

• Two persistent Betti numbers in dimension 1<br />

• No homology in dimension > 1<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

• Two persistent Betti numbers in dimension 1<br />

• No homology in dimension > 1<br />

• What is the space?<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

• Two persistent Betti numbers in dimension 1<br />

• No homology in dimension > 1<br />

• What is the space?<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Point-Cloud Data<br />

Persistence<br />

• LOTS of noise in dimension 0 at the beginning<br />

• Noise in dimension 1<br />

• One persistent Betti number in dimension 0<br />

• Two persistent Betti numbers in dimension 1<br />

• No homology in dimension > 1<br />

• What is the space?<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 30/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 31/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Kevin Knudson & Kenneth Brown<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Kevin Knudson & Kenneth Brown<br />

• Collection of speech signals from 3 individuals (cm, if, mc)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Kevin Knudson & Kenneth Brown<br />

• Collection of speech signals from 3 individuals (cm, if, mc)<br />

• Eight different sounds - aa, ae, eh, m, n, f, sh ,z<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Kevin Knudson & Kenneth Brown<br />

• Collection of speech signals from 3 individuals (cm, if, mc)<br />

• Eight different sounds - aa, ae, eh, m, n, f, sh ,z<br />

• For a given signal, let f (t) be the amplitude of the wave at<br />

time t<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Kevin Knudson & Kenneth Brown<br />

• Collection of speech signals from 3 individuals (cm, if, mc)<br />

• Eight different sounds - aa, ae, eh, m, n, f, sh ,z<br />

• For a given signal, let f (t) be the amplitude of the wave at<br />

time t<br />

• Form a Betti 0 barcode by “sweeping”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Kevin Knudson & Kenneth Brown<br />

• Collection of speech signals from 3 individuals (cm, if, mc)<br />

• Eight different sounds - aa, ae, eh, m, n, f, sh ,z<br />

• For a given signal, let f (t) be the amplitude of the wave at<br />

time t<br />

• Form a Betti 0 barcode by “sweeping”<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 32/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• There is a metric on barcodes that measures (dis)similarity<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• There is a metric on barcodes that measures (dis)similarity<br />

• The metric space this forms is called the barcode space<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• There is a metric on barcodes that measures (dis)similarity<br />

• The metric space this forms is called the barcode space<br />

• Each barcode corresponds <strong>to</strong> a point in the barcode space<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• There is a metric on barcodes that measures (dis)similarity<br />

• The metric space this forms is called the barcode space<br />

• Each barcode corresponds <strong>to</strong> a point in the barcode space<br />

• For each speaker, this gave a point cloud in the barcode<br />

space consisting of 800 points<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• There is a metric on barcodes that measures (dis)similarity<br />

• The metric space this forms is called the barcode space<br />

• Each barcode corresponds <strong>to</strong> a point in the barcode space<br />

• For each speaker, this gave a point cloud in the barcode<br />

space consisting of 800 points<br />

• They studied the structure of this point cloud, hoping there<br />

would be eight components, one for each sound<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 33/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Eight components for speakers cm and mc<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Eight components for speakers cm and mc<br />

• But for speaker if only obtain 3 components<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Eight components for speakers cm and mc<br />

• But for speaker if only obtain 3 components<br />

• Authors attribute this <strong>to</strong> noise (literally) in the data<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Eight components for speakers cm and mc<br />

• But for speaker if only obtain 3 components<br />

• Authors attribute this <strong>to</strong> noise (literally) in the data<br />

• Further research may lead <strong>to</strong> speech recognizition<br />

software or even speaker recognition software<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 34/ 42


Outline<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Simplicial Complex<br />

<strong>Homology</strong><br />

Applying <strong>Topology</strong><br />

Point-Cloud Data<br />

Persistence<br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 35/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Peter Bubenik, Peter Kim, et al.<br />

• Measured the cortical thickness of 16 autistic, 12 control<br />

right-handed, age-matched males<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 36/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Peter Bubenik, Peter Kim, et al.<br />

• Measured the cortical thickness of 16 autistic, 12 control<br />

right-handed, age-matched males<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 36/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 37/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 38/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 39/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Find the persistent homology of the thickness function<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 40/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Find the persistent homology of the thickness function<br />

• (Generalize the “sweeping” function above)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 40/ 42


<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Find the persistent homology of the thickness function<br />

• (Generalize the “sweeping” function above)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 40/ 42


Conclusion<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• It’s been common in the his<strong>to</strong>ry of mathematics for<br />

“useless” ideas <strong>to</strong> have applications after all (matrix theory,<br />

group theory, etc)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 41/ 42


Conclusion<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• It’s been common in the his<strong>to</strong>ry of mathematics for<br />

“useless” ideas <strong>to</strong> have applications after all (matrix theory,<br />

group theory, etc)<br />

• <strong>Algebraic</strong> <strong>to</strong>pology is being applied <strong>to</strong> sensor networks,<br />

statistics & data aggregation, robot motion, dynamics &<br />

numerical PDE’s, and other areas<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 41/ 42


Conclusion<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• It’s been common in the his<strong>to</strong>ry of mathematics for<br />

“useless” ideas <strong>to</strong> have applications after all (matrix theory,<br />

group theory, etc)<br />

• <strong>Algebraic</strong> <strong>to</strong>pology is being applied <strong>to</strong> sensor networks,<br />

statistics & data aggregation, robot motion, dynamics &<br />

numerical PDE’s, and other areas<br />

• <strong>Applied</strong> algebraic <strong>to</strong>pology is in its infancy - lots of<br />

problems (and grants!) are open (and available)<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 41/ 42


Conclusion<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• It’s been common in the his<strong>to</strong>ry of mathematics for<br />

“useless” ideas <strong>to</strong> have applications after all (matrix theory,<br />

group theory, etc)<br />

• <strong>Algebraic</strong> <strong>to</strong>pology is being applied <strong>to</strong> sensor networks,<br />

statistics & data aggregation, robot motion, dynamics &<br />

numerical PDE’s, and other areas<br />

• <strong>Applied</strong> algebraic <strong>to</strong>pology is in its infancy - lots of<br />

problems (and grants!) are open (and available)<br />

• Problem: Find better (more accurate) ways <strong>to</strong> build<br />

simplicial complexes<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 41/ 42


Conclusion<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• It’s been common in the his<strong>to</strong>ry of mathematics for<br />

“useless” ideas <strong>to</strong> have applications after all (matrix theory,<br />

group theory, etc)<br />

• <strong>Algebraic</strong> <strong>to</strong>pology is being applied <strong>to</strong> sensor networks,<br />

statistics & data aggregation, robot motion, dynamics &<br />

numerical PDE’s, and other areas<br />

• <strong>Applied</strong> algebraic <strong>to</strong>pology is in its infancy - lots of<br />

problems (and grants!) are open (and available)<br />

• Problem: Find better (more accurate) ways <strong>to</strong> build<br />

simplicial complexes<br />

• Problem: Statistical analysis<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 41/ 42


Thanks!<br />

<strong>Introduction</strong><br />

<strong>Topology</strong><br />

Applying <strong>Topology</strong><br />

Examples<br />

Phonemes<br />

Autism<br />

• Thank you for listening!<br />

DEPARTMENT OF MATHEMATICS, UNIVERSITY OF KENTUCKY<br />

URL: http://www.ms.uky.edu/∼jroberts<br />

E-mail address: jroberts@ms.uky.edu<br />

Joshua Roberts <strong>Applied</strong> <strong>Algebraic</strong> <strong>Topology</strong> 42/ 42

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!