Math Homotopy Theory Additional notes

Similar documents
For instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1)

Hermite Splines in Lie Groups as Products of Geodesics

b * -Open Sets in Bispaces

ALEKSANDROV URYSOHN COMPACTNESS CRITERION ON INTUITIONISTIC FUZZY S * STRUCTURE SPACE

Semi - - Connectedness in Bitopological Spaces

BASIC DIFFERENTIABLE MANIFOLDS AND APPLICATIONS MAPS WITH TRANSLATORS TANGENT AND COTANGENT SPACES

Complex Numbers. Now we also saw that if a and b were both positive then ab = a b. For a second let s forget that restriction and do the following.

CHAPTER 2 DECOMPOSITION OF GRAPHS

Some kinds of fuzzy connected and fuzzy continuous functions

Lecture 5: Multilayer Perceptrons

FLOW POLYTOPES OF PARTITIONS

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour

CMPS 10 Introduction to Computer Science Lecture Notes

Analysis of Continuous Beams in General

A NOTE ON FUZZY CLOSURE OF A FUZZY SET

On the diameter of random planar graphs

Ramsey numbers of cubes versus cliques

Life Tables (Times) Summary. Sample StatFolio: lifetable times.sgp

UNIT 2 : INEQUALITIES AND CONVEX SETS

Problem Set 3 Solutions

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique

Machine Learning. Support Vector Machines. (contains material adapted from talks by Constantin F. Aliferis & Ioannis Tsamardinos, and Martin Law)

Lecture 5: Probability Distributions. Random Variables

F Geometric Mean Graphs

PROPERTIES OF BIPOLAR FUZZY GRAPHS

such that is accepted of states in , where Finite Automata Lecture 2-1: Regular Languages be an FA. A string is the transition function,

TN348: Openlab Module - Colocalization

Accounting for the Use of Different Length Scale Factors in x, y and z Directions

Line geometry, according to the principles of Grassmann s theory of extensions. By E. Müller in Vienna.

Array transposition in CUDA shared memory

Clustering on antimatroids and convex geometries

Notes on Organizing Java Code: Packages, Visibility, and Scope

Data Representation in Digital Design, a Single Conversion Equation and a Formal Languages Approach

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

A New Approach For the Ranking of Fuzzy Sets With Different Heights

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification

Mathematics 256 a course in differential equations for engineering students

An Approach in Coloring Semi-Regular Tilings on the Hyperbolic Plane

Harmonic Coordinates for Character Articulation PIXAR

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Intro. Iterators. 1. Access

Any Pair of 2D Curves Is Consistent with a 3D Symmetric Interpretation

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics

the nber of vertces n the graph. spannng tree T beng part of a par of maxmally dstant trees s called extremal. Extremal trees are useful n the mxed an

Introduction. Leslie Lamports Time, Clocks & the Ordering of Events in a Distributed System. Overview. Introduction Concepts: Time


Performance Evaluation of Information Retrieval Systems

Cordial and 3-Equitable Labeling for Some Star Related Graphs

A Five-Point Subdivision Scheme with Two Parameters and a Four-Point Shape-Preserving Scheme

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

ON SOME ENTERTAINING APPLICATIONS OF THE CONCEPT OF SET IN COMPUTER SCIENCE COURSE

Rough Neutrosophic Multisets Relation with Application in Marketing Strategy

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements

Optimization Methods: Integer Programming Integer Linear Programming 1. Module 7 Lecture Notes 1. Integer Linear Programming

On Some Entertaining Applications of the Concept of Set in Computer Science Course

Detection of an Object by using Principal Component Analysis

Harvard University CS 101 Fall 2005, Shimon Schocken. Assembler. Elements of Computing Systems 1 Assembler (Ch. 6)

Mercer Kernels for Object Recognition with Local Features

Computers and Mathematics with Applications. Discrete schemes for Gaussian curvature and their convergence

EXTENDED FORMAL SPECIFICATIONS OF 3D SPATIAL DATA TYPES

On Embedding and NP-Complete Problems of Equitable Labelings

5 The Primal-Dual Method

Reactors: A Data-Oriented Synchronous/Asynchronous Programming Model for Distributed Applications

LOOP ANALYSIS. The second systematic technique to determine all currents and voltages in a circuit

GSLM Operations Research II Fall 13/14

Concurrent models of computation for embedded software

Non-Split Restrained Dominating Set of an Interval Graph Using an Algorithm

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

Constructing Minimum Connected Dominating Set: Algorithmic approach

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Active Contours/Snakes

LECTURE : MANIFOLD LEARNING

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data

Bridges and cut-vertices of Intuitionistic Fuzzy Graph Structure

Fitting: Deformable contours April 26 th, 2018

Secure Index Coding: Existence and Construction

Solving some Combinatorial Problems in grid n-ogons

Oracle Database: SQL and PL/SQL Fundamentals Certification Course

A Note on Quasi-coincidence for Fuzzy Points of Fuzzy Topology on the Basis of Reference Function

NETWORKS of dynamical systems appear in a variety of

Fuzzy soft -ring. E Blok Esenler, Istanbul, Turkey 2 Department of Mathematics, Marmara University, Istanbul, Turkey

On 3D DDFV discretization of gradient and divergence operators. I. Meshing, operators and discrete duality.

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices

Cell Count Method on a Network with SANET

Lecture 4: Principal components

Universal Hinge Patterns for Folding Orthogonal Shapes

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

(1) The control processes are too complex to analyze by conventional quantitative techniques.

Assembler. Shimon Schocken. Spring Elements of Computing Systems 1 Assembler (Ch. 6) Compiler. abstract interface.

Electrical analysis of light-weight, triangular weave reflector antennas

AP PHYSICS B 2008 SCORING GUIDELINES

T. Background material: Topology

S.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION?

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

Visual Curvature. 1. Introduction. y C. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2007

Stable Structures of Coalitions in Competitive and Altruistic Military Teams. M. Aurangzeb, D. Mikulski, G, Hudas, F.L. Lewis, and Edward Gu

Specifying Database Updates Using A Subschema

A Unified, Integral Construction For Coordinates Over Closed Curves

Secure Index Coding: Existence and Construction

treated this question with techniques permitting proofs. A synthesis of their

Transcription:

Math 527 - Homotopy Theory Addtonal notes Martn Frankland February 4, 2013 The category Top s not Cartesan closed. problem. In these notes, we explan how to remedy that 1 Compactly generated spaces Ths secton and the next are essentally taken from [3, 1,2]. 1.1 Basc defntons and propertes Defnton 1.1. Let X be a topologcal space. A subset A X s called k-closed n X f for any compact Hausdorff space K and contnuous map u: K X, the premage u 1 (A) K s closed n K. The collecton of k-closed subsets of X forms a topology, whch contans the orgnal topology of X (.e. closed subsets are always k-closed). Notaton 1.2. Let kx denote the space whose underlyng set s that of X, but equpped wth the topology of k-closed subsets of X. Because the k-topology contans the orgnal topology on X, the dentty functon d: kx X s contnuous. Defnton 1.3. A space X s compactly generated (CG), sometmes called a k-space, f kx X s a homeomorphsm. In other words, every k-closed subset of X s closed n X. Example 1.4. Every locally compact space s CG. Example 1.5. Every frst-countable space s CG. More generally, every sequental space s CG. Example 1.6. Every CW-complex s CG. Notaton 1.7. Let CG denote the full subcategory of Top consstng of compactly generated spaces. Notaton 1.8. The constructon of kx defnes a functor k : Top CG, called the k-fcaton functor. Proposton 1.9. Let X be a CG space and Y an arbtrary space. Then a functon f : X Y s contnuous f and only f for every compact Hausdorff space K and contnuous map u: K X, the composte fu : K Y s contnuous. 1

Proposton 1.10. For any space X, we have k 2 X = kx, so that kx s always compactly generated. Proposton 1.11. Let X be a CG space and Y an arbtrary space. Then a functon f : X Y s contnuous f and only t s contnuous when vewed as a functon f : X ky. Proposton 1.11 can be reformulated n the followng more suggestve way, as a unversal property. For any CG space X and contnuous map f : X Y, there exsts a unque contnuous map f : X ky satsfyng f = d f,.e. makng the dagram f ky d Y X f commute. Note that f has the same underlyng functon as f. Ths exhbts ky Y as the closest approxmaton of Y by a CG space. Corollary 1.12. The k-fcaton functor k : Top CG s rght adjont to the ncluson ι: CG Top. In other words, CG s a coreflectve subcategory of Top. The dentty functon ιkx X s the count of the adjuncton, whereas the unt W kιw s the dentty map for any CG space W. Proposton 1.13. 1. The category CG s complete. Lmts n CG are obtaned by applyng k to the lmt n Top. 2. The category CG s cocomplete. Colmts n CG are computed n Top. Proof. Let I be a small category and F : I CG an I-dagram. Let us wrte X := F () and, by abuse of notaton, lm X := lm I F. (1) Vewng the CG spaces X as spaces ιx, we can compute the lmt of the dagram ιf snce Top s complete (c.f. Homework 4 Problem 2). Applyng k yelds the CG space k(lm ιx ). For any CG space W, we have a natural somorphsm Hom CG (W, k(lm ιx )) = Hom Top (ιw, lm ιx ) = lm Hom Top (ιw, ιx ) = lm Hom CG (W, X ) where the last equalty comes from the fact that CG s a full subcategory of Top. Ths proves k(lm ιx ) = lm X. (2) We can compute the colmt X = colm ιx of the dagram ιf snce Top s cocomplete (c.f. Homework 4 Problem 2 and Remark afterwards). Snce X s a quotent of a coproduct of CG spaces ιx, X s also CG, by [3, Prop. 2.1, Prop. 2.2]. Moreover t s the desred colmt 2

n CG. For any CG space Y, we have a natural somorphsm Hom CG (X, Y ) = Hom Top (ιx, ιy ) = Hom Top (ι colm ιx, ιy ) = Hom Top (colm ιx, ιy ) = lm Hom Top (ιx, ιy ) whch proves X = colm X. = lm Hom CG (X, Y ) In partcular, products n CG may not agree wth the usual product n Top. Notaton 1.14. For CG spaces X and Y, wrte X 0 Y = ιx ιy for ther usual product n Top, and wrte X Y = k(x 0 Y ) for ther product n CG. 1.2 Mappng spaces Defnton 1.15. Let X and Y be CG spaces. For any compact Hausdorff space K, contnuous map u: K X, and open subset U Y, consder the set W (u, K, U) := {f : X Y contnuous fu(k) U}. Denote by C 0 (X, Y ) the set of contnuous maps from X to Y, equpped wth the topology generated by all such subsets W (u, K, U). Ths topology s called the compact-open topology. Note that C 0 (X, Y ) need not be CG. Wrte Map(X, Y ) := kc 0 (X, Y ). Theorem 1.16. For any CG spaces X, Y, and Z, the natural map s a homeomorphsm. ϕ: Map(X Y, X) Map(X, Map(Y, Z)) (1) The fact that ϕ s bjectve tells us that CG s Cartesan closed, n the unenrched sense. The theorem s even better: CG s Cartesan closed, n the enrched sense. Note that CG s enrched n tself, gven that the composton map s contnuous. Map(X, Y ) Map(Y, Z) Map(X, Z) Remark 1.17. The exponental object Map(X, Y ) s often denoted Y X. The somorphsm (1), whch can be wrtten as Z X Y = (Z Y ) X s often called the exponental law. 3

2 Weakly Hausdorff spaces The category CG would be good enough to work wth, but we can also mpose a separaton axom to our spaces. Defnton 2.1. A topologcal space X s weakly Hausdorff (WH) f for every compact Hausdorff space K and every contnuous map u: K X, the mage u(k) X s closed n X. Remark 2.2. Hausdorff spaces are weakly Hausdorff, snce u(k) s compact and thus closed n X f X s Hausdorff. Ths justfes the termnology. Moreover, weakly Hausdorff spaces are T 1, snce the sngle pont space s compact Hausdorff. Thus we have mplcatons Hausdorff weakly Hausdorff T 1. Example 2.3. Every CW-complex s Hausdorff, hence n partcular WH. Proposton 2.4. If X s a WH space, then any larger topology on X s stll WH. In partcular, kx s stll WH. Proof. Let X be the set X equpped wth a topology contanng the orgnal topology,.e. the dentty functon d: X X s contnuous. For any compact Hausdorff space K and contnuous map u: K X, the composte du: K X s contnuous and so ts mage du(k) X s closed n X. Thus u(k) = d 1 du(k) s closed n X. Proposton 2.5. Any subspace of a WH space s WH. Proof. Let X be a WH space and : A X the ncluson of a subspace. For any compact Hausdorff space K and contnuous map u: K A, the composte u: K X s contnuous and so ts mage u(k) X s closed n X, and thus n A as well. Notaton 2.6. Let CGWH denote the full subcategory of CG consstng of compactly generated weakly Hausdorff spaces. Defnton 2.7. For any CG space X, let hx be the quotent of X by the smallest closed equvalence relaton on X (see [3, Prop. 2.22]). Then hx s stll CG snce t s a quotent of a CG space [3, Prop. 2.1], and t s WH snce we quotented out a closed equvalence relaton on X [3, Cor. 2.21]. Ths defnes a functor h: CG CGWH called weak Hausdorfffcaton By constructon, the quotent map q : X hx satsfes the followng unversal property. For any CGWH space Y and contnuous map f : X Y, there exsts a unque contnuous map f : hx Y satsfyng f = fq,.e. makng the dagram X q hx f f Y commute. Ths exhbts X hx as the closest approxmaton of X by a CGWH space. 4

Corollary 2.8. The functor h: CG CGWH s left adjont to the ncluson functor ι: CG CGWH. In other words, CGWH s a reflectve subcategory of CG. The quotent map q : X hx s the unt of the adjuncton, whereas the count hιw W s the dentty map for any CGWH space W. Proposton 2.9. n CG. 1. The category CGWH s complete. Lmts n CGWH are computed 2. The category CGWH s cocomplete. Colmts n CGWH are obtaned by applyng h to the colmt n CG. Proof. Let I be a small category and F : I CGWH an I-dagram. (1) The lmt X = lm ιx computed n CG, whch exsts snce CG s complete (by 1.13), s stll WH. Indeed, an arbtrary product n CG of CGWH spaces s stll WH [3, Cor. 2.16], and so s an equalzer n CG of two maps (by 2.5 and 2.4). Therefore X s also the lmt n CGWH, by the same argument as 1.13 (2). (2) We have h(colm ιx ) = colm X n CGWH by the same argument as 1.13 (1). To summarze the stuaton, there are two adjont pars as follows: ι CG Top k h ι CGWH Proposton 2.10. If X s a CG space and Y s a CGWH space, then Map(X, Y ) s CGWH. Consequently, the category CGWH s enrched n tself. Note that t s also Cartesan closed (n the enrched sense). Indeed, for any X, Y, and Z n CGWH, the natural map s a homeomorphsm. ϕ: Map(X Y, X) Map(X, Map(Y, Z)) 5

3 A convenent category of spaces In ths secton, we explan n what sense t s preferable to work wth the category CGWH nstead of Top. We follow the treatment n [1], tself nspred by [2]. Defnton 3.1. A convenent category of topologcal spaces s a full replete (meanng closed under somorphsms of objects) subcategory C of Top satsfyng the followng condtons. 1. All CW-complexes are objects of C. 2. C s complete and cocomplete. 3. C s Cartesan closed. Note that CG and CGWH are replete full subcategores of Top, snce both condtons of beng CG or WH are nvarant under homeomorphsm. Let us summarze the dscusson as follows. Proposton 3.2. The categores CG and CGWH are convenent. In fact, there are other desrable propertes for a convenent category of spaces. For nstance, one would lke that closed subspaces of objects n C also be n C. Both CG and CGWH satsfy ths addtonal condton. Proposton 3.3. Consder nclusons of spaces A B X. If A s k-closed n B and B s k-closed n X, then A s k-closed n X. Proof. Let K be a compact Hausdorff space and u: K X a contnuous map. Then the premage u 1 (B) s closed n K, hence compact (and also Hausdorff). Consder the restrcton u u 1 (B) : u 1 (B) B. Snce A s k-closed n B, the premage u 1 u 1 (B) (A) = u 1 (A) s closed n u 1 (B) and thus n K as well. Corollary 3.4. A closed subspace of a CG space s also CG. Proof. Let X be a CG space and B X a closed subspace. Let A B be a k-closed subset of B. Then A s k-closed n X (snce B s closed n X), hence closed n X (snce X s CG). Therefore A s closed n B. Corollary 3.5. A closed subspace of a CGWH space s also CGWH. 6

4 Some applcatons Here s a toy example llustratng the use of the exponental law. Proposton 4.1. For any spaces X and Y, the natural map [X, Y ] bjecton. Proof. Snce the map Map(X I, Y ) Map(I, Map(X, Y )) = π 0 Map(X, Y ) s a s a bjecton, two maps f, g : X Y are homotopc f and only f they are connected by a (contnuous) path n Map(X, Y ). Here s another beneft of workng n the category CG. Proposton 4.2. If X and Y are CW-complexes, then X Y nherts a CW-structure, where a p-cell of X and a q-cell of Y produce a (p + q)-cell of X Y. Here we do mean the product X Y s CG, not n Top. A pror, the product X 0 Y n Top could fal to be a CW-complex. See Hatcher (A.6) for detals. References [1] Ronald Brown, nlab: Convenent category of topologcal spaces (2012), avalable at http://ncatlab. org/nlab/show/convenent+category+of+topologcal+spaces. [2] N. E. Steenrod, A convenent category of topologcal spaces, Mchgan Math. J. 14 (1967), 133 152. [3] Nel Strckland, The category of CGWH spaces (2009), avalable at http://www.nel-strckland.staff. shef.ac.uk/courses/homotopy/cgwh.pdf. 7