Fact-checked by Grok 2 weeks ago

H-space

In algebraic topology, an H-space is a pointed topological space (X, e) equipped with a continuous multiplication map \mu: X \times X \to X such that the restrictions \mu|_{ \{e\} \times X } and \mu|_{ X \times \{e\} } are each homotopic to the corresponding projections onto X, and \mu is associative up to homotopy (i.e., there exists a homotopy between \mu \circ (\mu \times \mathrm{id}_X) and \mu \circ (\mathrm{id}_X \times \mu)).^[1] This structure generalizes topological groups by relaxing strict associativity and unit conditions to their homotopy-theoretic counterparts, allowing the space to behave like a group in homotopy theory while permitting more flexible examples.^[1] The concept was introduced by Heinz Hopf in 1941 to study the homology of group manifolds and their generalizations, where he defined an H-space as a space with a continuous multiplication admitting a two-sided unit. H-spaces play a central role in homotopy theory, as their multiplication induces compatible algebraic structures on homotopy groups and homology. For instance, the multiplication on an H-space X defines a group operation on each homotopy group \pi_n(X, e) for n \geq 1, making \pi_n(X, e) abelian for n \geq 2 and trivializing the action of \pi_1(X, e) on higher homotopy groups.^[1] In homology, the map \mu yields a Pontryagin product, turning H_*(X; R) into a graded algebra for a commutative ring R, and under suitable conditions (e.g., path-connectedness), the cohomology H^*(X; R) forms a Hopf algebra.^[1] If the H-space is homotopy associative and commutative with a strict identity, its homotopy type is determined by its homotopy groups via a weak equivalence to a product of Eilenberg–MacLane spaces K(G_n, n).^[1] Notable examples of H-spaces include topological groups like the circle S^1, the general linear groups GL_n(\mathbb{R}), GL_n(\mathbb{C}), and GL_n(\mathbb{H}), where the group operation provides the multiplication.^[1] Among spheres, only S^0, S^1, S^3, and S^7 admit H-space structures, with the multiplications on S^3 and S^7 arising from the normed division algebras \mathbb{H} and the octonions, respectively; these are tied to Hopf's original investigations into Hopf invariants.^[1] Loop spaces \Omega X form H-spaces under composition of loops, and infinite projective spaces like \mathbb{CP}^\infty and \mathbb{RP}^\infty are associative and commutative H-spaces.^[1] More broadly, the loop space of a suspension \Omega \Sigma X is homotopy equivalent to the James construction J(X), an infinite-dimensional H-space, highlighting connections to stable homotopy theory.^[1]

Definition and Basic Concepts

Formal Definition

An H-space is a pointed topological space (X, e), where e \in X is the basepoint, together with a continuous multiplication map \mu: X \times X \to X satisfying \mu(e, e) = e. The basepoint e acts as a homotopy unit, meaning the left unit map \lambda: X \to X defined by \lambda(x) = \mu(x, e) and the right unit map \rho: X \to X defined by \rho(x) = \mu(e, x) are both homotopic to the identity map \mathrm{id}_X, with these homotopies holding relative to the basepoint e.^[2]^[3] Unlike a topological group, which requires a strictly associative multiplication and continuous inverses, an H-space does not demand homotopy associativity or the existence of inverses; the multiplication \mu need only satisfy the homotopy unit condition. This generalization allows H-spaces to capture algebraic structures that are invariant under homotopy equivalence, such as loop spaces.^[4] A strict H-space is a special case where the unit maps \lambda and \rho are exactly equal to the identity map \mathrm{id}_X, without needing non-constant homotopies. In this setting, the basepoint serves as a strict two-sided unit for the multiplication.^[3]^[2]

Equivalent Characterizations

For a CW-complex X with basepoint e as a 0-cell, several variants of the H-space definition are equivalent: one requiring the homotopy units to fix the basepoint throughout the homotopy, another allowing homotopies that do not fix the basepoint, and a third requiring a strict unit without homotopy. This equivalence relies on the cellular structure of CW-complexes, which permits approximations that preserve the homotopy-theoretic unit condition.^[1] When every element in an H-space admits a homotopy inverse, the space becomes an H-group, which is a monoid object in the homotopy category of pointed topological spaces equipped with inverses up to homotopy. In this setting, the H-group acts as a group object, enabling deloopings and connections to infinite loop spaces under suitable conditions. An H-group requires the multiplication to be homotopy associative in addition to having homotopy inverses.^[1]^[5] Unpointed H-spaces can be characterized without a distinguished basepoint, where the multiplication map admits left and right units up to homotopy, meaning there exist maps homotopic to the projections onto one factor such that composing with the multiplication yields the identity on the space. This formulation extends the pointed case to the unpointed homotopy category, preserving the essential algebraic structure up to homotopy.^[5]

Examples

Division Algebra Structures on Spheres

Normed division algebras over the real numbers induce H-space structures on their unit spheres through the restriction of the algebra's multiplication map, normalized to preserve the unit sphere.^[6] By Hurwitz's theorem, the only such finite-dimensional algebras are the reals (dimension 1), complexes (dimension 2), quaternions (dimension 4), and octonions (dimension 8), up to isomorphism.^[7] These yield H-space multiplications precisely on the spheres S^0, S^1, S^3, and S^7, respectively.^[8] The 0-sphere S^0, consisting of the two points \{1, -1\} as the unit sphere in \mathbb{R}, acquires an H-space structure via real multiplication, which makes it isomorphic to the discrete group \mathbb{Z}/2\mathbb{Z}.^[8] Similarly, the 1-sphere S^1 is the unit circle in \mathbb{C}, and complex multiplication endows it with the structure of the Lie group U(1), which is associative and commutative.^[8] For the 3-sphere S^3, identified with the unit quaternions in \mathbb{H}, quaternion multiplication provides an associative H-space structure, realizing S^3 as the Lie group SU(2).^[8] In each of these cases, the multiplication is strictly associative, forming topological groups. The 7-sphere S^7, corresponding to the unit octonions in \mathbb{O}, receives a non-associative multiplication from the octonion algebra, yet this defines an H-space structure because the multiplication is homotopy associative.^[8] Specifically, the associator (xy)z - x(yz) is nullhomotopic when restricted to the unit sphere, allowing a homotopy between left and right multiplication that establishes the required homotopy associativity.^[6] However, unlike the lower-dimensional cases, this H-space on S^7 does not extend to an A_\infty-space, as higher coherences fail.^[5] Adams' theorem confirms the exclusivity of these structures: up to homotopy equivalence, the only spheres admitting H-space multiplications are S^0, S^1, S^3, and S^7, corresponding to the existence of maps S^{2n-1} \to S^n of Hopf invariant one, which are realized by the Hopf fibrations associated to these division algebras.^[9] This result, proved using secondary cohomology operations, rules out H-space structures on other spheres.^[9]

Loop Spaces and Free Loop Spaces

The based loop space \Omega X of a pointed topological space (X, x_0) consists of all continuous maps \gamma: [0,1] \to X such that \gamma(0) = \gamma(1) = x_0, equipped with the compact-open topology. This space admits a natural H-space structure via the concatenation product \mu: \Omega X \times \Omega X \to \Omega X, defined for \alpha, \beta \in \Omega X by

(\alpha * \beta)(t) = \begin{cases} \alpha(2t) & 0 \leq t \leq 1/2, \\ \beta(2t - 1) & 1/2 \leq t \leq 1. \end{cases}

The constant loop at x_0 serves as the identity element, and the multiplication \mu is continuous, rendering \Omega X an H-space for any pointed space X.^[1] More specifically, the loop space of a suspension, \Omega \Sigma X, is homotopy equivalent to the James construction J(X), which is an infinite-dimensional free H-space model for it, highlighting its connections to stable homotopy theory.^[1] The reversal map \iota: \Omega X \to \Omega X, given by \iota(\alpha)(t) = \alpha(1 - t), provides a homotopy inverse to the identity with respect to the concatenation product, as \mu(\alpha, \iota(\alpha)) and \mu(\iota(\alpha), \alpha) are both homotopic to the constant loop. This structure makes \Omega X an H-group, a special class of H-spaces where homotopy inverses exist. When X is simply connected, \Omega X is grouplike, meaning the monoid of path components under the induced multiplication is a group, and the H-space multiplication is homotopy associative in a coherent manner related to its delooping properties.^[1]^[8] The free loop space LX = \mathrm{Map}(S^1, X) comprises all continuous maps from the circle S^1 to X, again with the compact-open topology. If X itself is an H-space with multiplication m: X \times X \to X, then LX inherits an H-space structure via pointwise multiplication: for f, g \in LX, define (f \cdot g)(\theta) = m(f(\theta), g(\theta)) for all \theta \in S^1. The constant map to the identity of X acts as the unit. In cases where X is simply connected, this pointwise H-space on LX aligns with the delooping of related structures, such as when X serves as a classifying space, though the components of LX may not form a single connected H-space.

Topological Groups as H-spaces

A topological group G equips a group structure with a topology such that the multiplication map G \times G \to G and the inversion map G \to G are continuous, rendering G an H-space with the identity element e serving as the strict two-sided unit.^[3] The continuous multiplication directly provides the required H-space multiplication, and the strict unit satisfies the homotopy unit condition without needing deformation.^[10] Unlike general H-spaces, where multiplication is associative and unital only up to homotopy, topological groups feature strictly associative multiplication and strict two-sided inverses for every element, imposing a rigid algebraic structure on the topological framework.^[3] This additional rigidity makes topological groups prototypical examples in homotopy theory, as their operations align precisely without homotopy corrections.^[10] Prominent compact examples include Lie groups such as the special orthogonal group SO(n), which consists of rotations in \mathbb{R}^n, and the unitary group U(n), comprising unitary matrices, both inheriting H-space structures from their continuous group operations.^[3] These compact Lie groups illustrate how finite-dimensional smooth manifolds can carry compatible group topologies, facilitating applications in representation theory and geometry.

Classifying Spaces and Infinite Projective Spaces

Infinite projective spaces provide additional examples of H-spaces. The complex projective space \mathbb{CP}^\infty, which is the classifying space BU(1) for complex line bundles, admits an associative and commutative H-space structure induced by the tensor product of line bundles. Similarly, the real projective space \mathbb{RP}^\infty = BO(1) is an H-space via the operation corresponding to the direct sum of line bundles (or symmetric product). These structures make them models for Eilenberg–MacLane spaces K(\mathbb{Z}, 2) and K(\mathbb{Z}/2\mathbb{Z}, 1), respectively, with compatible algebraic operations in homotopy theory.^[1] Although all topological groups qualify as H-spaces, infinite discrete groups lacking compact generation do not typically serve as meaningful examples in homotopy-theoretic contexts, where the standard category emphasizes compactly generated spaces to ensure well-behaved limits and colimits.^[3]

Algebraic and Homotopy Properties

Properties of Homotopy Groups

In path-connected H-spaces, the fundamental group \pi_1(X, e) is abelian. This follows from the Eckmann-Hilton argument, which shows that the H-space multiplication induces a commutative binary operation on homotopy classes of based maps from the circle S^1 to X, identified with \pi_1(X, e).^[11] To sketch the proof, consider two loops \gamma, \delta: I \to X based at the identity e. The commutator [\gamma, \delta] is the loop formed by composing \gamma followed by \delta, then the inverse of \gamma, and the inverse of \delta. Using the H-space multiplication \mu: X \times X \to X, one constructs a homotopy H: I \times I \to X that deforms the commutator map to the constant loop at e, by leveraging the homotopy units and the continuous multiplication to interchange the paths continuously. This homotopy implies [\gamma, \delta] \simeq [\delta, \gamma] in \pi_1(X, e), establishing commutativity and thus abelianity.^[1]^[12] The H-space multiplication \mu further induces bilinear operations on the higher homotopy groups. For n \geq 2, \pi_n(X, e) is already an abelian group under the standard pointwise addition of maps S^n \to X, and \mu_* provides an additional group structure via (\alpha \cdot \beta)(p) = \mu(\alpha(p), \beta(p)) for \alpha, \beta \in \pi_n(X, e) and p \in S^n, compatible with the existing addition up to homotopy. This endows the collection \{\pi_n(X, e)\}_{n \geq 1} with a graded ring structure, where \pi_1(X, e) acts on higher groups.^[1] A key operation arising from the H-space structure is the Samelson product, a bilinear map \langle -, - \rangle: \pi_n(X, e) \times \pi_m(X, e) \to \pi_{n+m-1}(X, e) defined for n, m \geq 1. It generalizes the Whitehead product and captures the failure of commutativity in the H-space multiplication, constructed by composing representatives with the commutator [\mu, \Delta] (where \Delta is the diagonal) up to homotopy, yielding an element in the homotopy group of one lower dimension. The Samelson product satisfies graded Lie algebra identities up to sign and vanishes when the H-space multiplication is homotopy commutative.^[13]^[1]

Cohomology and Homology Structures

In an H-space (X, \mu), where \mu: X \times X \to X is the multiplication map, the cohomology ring H^*(X; R) acquires a natural Hopf algebra structure over a commutative ring R, with the coproduct \Delta: H^*(X; R) \to H^*(X; R) \otimes_R H^*(X; R) induced by the pullback \mu^*. This coproduct satisfies \Delta(\alpha) = \mu^*(\alpha) for \alpha \in H^*(X; R), and the unit and counit are determined by the homotopy unit map, ensuring the bialgebra axioms hold up to homotopy, which become strict in cohomology.^[14] For a path-connected H-space X whose cohomology H^*(X; R) is free and finitely generated as an R-module, the resulting structure is a Hopf algebra, complete with an antipode arising from the homotopy inverse when available. This finiteness condition ensures the existence of the antipode, as graded connected bialgebras over such rings automatically possess one, mirroring the algebraic dual to the group-like structure in homotopy. Seminal results, such as generalizations of Hopf's theorem, confirm that such cohomology rings are exterior algebras on odd-degree generators when R is a field of characteristic zero.^[14] Dually, the homology groups H_*(X; R) support the Pontryagin product, defined as the pushforward \mu_*: H_*(X \times X; R) \to H_*(X; R), which endows H_*(X; R) with a ring structure compatible with the intersection product on X \times X. This product is associative up to homotopy and serves as the algebraic dual to the cohomology coproduct, establishing a pairing that interchanges the two operations; specifically, the Pontryagin product on homology corresponds to the coproduct on cohomology via cap or duality pairings when applicable. For simply connected H-spaces, this yields a Hopf algebra on homology as well, with the diagonal map providing the coproduct.^[15] When coefficients are taken in a prime field \mathbb{F}_p, the actions of the Steenrod algebra \mathcal{A}_p on both H^*(X; \mathbb{F}_p) and H_*(X; \mathbb{F}_p) are compatible with their Hopf structures, meaning the coproduct and product respect the module algebra and comodule coalgebra actions. The Steenrod operations, such as squares or powers, commute with the Hopf operations in the sense that \Delta \circ Sq^k = (Sq^k \otimes Sq^k) \circ \Delta (and dually for homology), preserving the bialgebra axioms and enabling deeper invariants like the Milnor-Moore theorem for decompositions into primitives. This compatibility has been pivotal in classifying finite H-spaces modulo p.^[15]^[16]

Historical Development and Applications

Origins and Key Contributions

The concept of H-spaces emerged from early investigations into the topological structure of spheres and their mappings, with foundational contributions tracing back to Heinz Hopf's work in the 1930s. In his 1931 paper, Hopf introduced the notion of fibrations of spheres and defined a homotopy invariant now known as the Hopf invariant, which quantifies the linking of fibers in maps from odd-dimensional spheres to even-dimensional ones. This invariant played a crucial role in identifying multiplications on spheres, particularly those arising from complex and quaternionic structures, laying the groundwork for understanding spaces with continuous binary operations up to homotopy. Hopf further developed these ideas in his 1941 paper "Über die Topologie der Gruppen-Manigfaltigkeiten und ihrer Verallgemeinerungen", where he defined spaces equipped with a continuous binary operation admitting a two-sided unit, generalizing the structure of Lie groups to topological settings.^[17] The term "H-space" was coined by Jean-Pierre Serre in the 1950s, specifically in his 1951 paper on the singular homology of fiber spaces, where he used "H" to honor Hopf's influence on the subject.^[18] Serre formalized H-spaces as topological spaces equipped with a continuous multiplication and a unit element, associative and unital up to homotopy, and applied this framework to study the homotopy groups of spheres through fibrations and loop spaces.^[19] His work demonstrated that certain spheres, such as those of dimensions 1, 3, and 7, admit H-space structures derived from division algebras, providing early examples that highlighted the concept's relevance to algebraic topology.^[18] A pivotal advancement came in 1960 with J. Frank Adams' resolution of the Hopf invariant one problem, proving that maps S^{2n-1} → S^n of Hopf invariant one exist only for n=2,4,8, thereby determining that only the spheres S^1, S^3, and S^7 admit H-space multiplications. Adams employed the newly developed Adams spectral sequence to establish this theorem, which had profound implications for classifying division algebras and exotic structures on spheres.^[9] The 1950s and 1960s saw H-spaces evolve alongside stable homotopy theory, with researchers like Samuel Eilenberg and others integrating these structures into broader studies of homotopy categories and spectral sequences.^[8]

Applications in Homotopy Theory

H-spaces are fundamental in the delooping construction within homotopy theory, where a grouplike H-space—characterized by a homotopy inverse and multiplication inducing an abelian monoid on homotopy classes—admits a delooping \mathbf{B}X such that the loop space \Omega(\mathbf{B}X) \simeq X up to homotopy equivalence in the homotopy category of topological spaces.^[20] This correspondence identifies H-spaces with connected components of the Picard groupoid of the stable homotopy category, enabling the construction of classifying spaces for principal bundles and fibrations.^[21] For abelian groups G, the Eilenberg-MacLane space K(G, n) serves as the delooping of the loop space \Omega K(G, n+1), which carries a natural H-space structure, facilitating the study of spaces with a single nontrivial homotopy group.^[20] H-spaces represent a strict case of A_\infty-spaces, where the multiplication satisfies associativity up to coherent higher homotopies rather than strictly, allowing for more flexible structures in the (\infty,1)-category of spaces.^[5] The loop space \Omega X of any pointed space X is canonically an A_\infty-space, inheriting an H-space multiplication from concatenation of loops, but with higher coherences that resolve homotopy inconsistencies in non-associative cases.^[5] This relation extends to modern frameworks, where E_\infty-spaces generalize H-spaces further by incorporating \Sigma_\infty-equivariant structures via operads like the little cubes operad, enabling the recognition of infinite loop spaces and the delooping to spectra representing generalized cohomology theories.^[21] In computing homotopy groups, H-spaces simplify the analysis of Postnikov towers by imposing algebraic constraints on the tower's stages; for a connected space, the Postnikov invariant in the cohomology group H^3(π_1(X); π_2(X)) must vanish (or be additive) for the space to admit an H-space structure up to 2-type.^[22] This allows approximations of homotopy types through H-space models, where the tower's fibers are Eilenberg-MacLane spaces, and the multiplication induces compatible group structures on homotopy groups, aiding explicit calculations via spectral sequences.^[23] H-spaces connect deeply to topological K-theory, a generalized cohomology theory where the classifying space BU for stable complex vector bundles is a grouplike H-space under Whitney sum, with the spectrum BU \times \mathbb{Z} representing K-theory and enabling the classification of bundles via maps to BU.^[24] This structure extends to other theories, such as real K-theory via BO, where H-space multiplications correspond to tensor products of bundles, providing multiplicative refinements to cohomology and tools like Adams operations for detecting obstructions in homotopy groups.^[25] Contemporary applications leverage Goodwillie calculus to approximate functors on H-spaces, decomposing the identity functor or embedding functors into Taylor towers whose layers reveal unstable homotopy information; for instance, this yields computations of \nu_1-periodic homotopy groups of simply connected finite H-spaces by analyzing polynomial approximations and their connectivity.^[26] Such techniques highlight H-spaces' role in functorial homotopy theory, bridging classical structures to higher categorical insights without relying on exhaustive group listings.^[27]

References

[1]
[PDF] Algebraic Topology - Cornell Mathematics
One of the main ideas of algebraic topology is to consider two spaces to be equiv- alent if they have 'the same shape' in a sense that is much broader than ...
[2]
[PDF] Homotopy Associativity of H-Spaces. I
The concept of an H-space arose as a generalization of that of a topological group. The essential feature which is retained is a continuous multiplication with ...
[3]
[PDF] MATH 6280 - CLASS 5 Contents 1. H–Spaces 2 2. H–cospaces 6 ...
Definition 1.4. An H-space W is a pointed topological space and a continuous map. µ : W × W → W such that the map e : W → W, e(w) = ∗ is a unit up to homotopy.
[4]
[PDF] A Concise Course in Algebraic Topology J. P. May
By definition, a space X is Hausdorff if the diagonal subspace ∆X ... An “H-space” is a space X with a basepoint e and a product φ : X × X −→. X ...
[5]
[PDF] REVIEW ON HIGHER HOMOTOPIES IN THE THEORY OF H-SPACES
The other version is that the base point is only assumed to be a homotopy unit, that is, both maps x 7→ µX(x,∗) and x 7→ µX(∗,x) are homotopic to the identity ...
[6]
H-space in nLab
Jun 8, 2025 · 1. Definition. An H -space (“H” for Hopf, as in Hopf construction) is a magma internal to the classical homotopy category of topological spaces ...
[7]
normed division algebra in nLab
Aug 21, 2024 · Normed division algebras · Classification · Automorphisms · Relation to H-space structures on sphere (Hopf invariant one) · Magic square.
[8]
Hurwitz theorem in nLab
Oct 2, 2025 · The Hurwitz Theorem can be extended to more general composition algebras that are not necessarily unital, but that are para-unital or even non-unital.
[9]
[PDF] H-Spaces I
Jun 29, 2020 · The concept of an H-space is what results when that of a topological magma is generalised to a homotopy-invariant notion.
[10]
[PDF] On the Non-Existence of Elements of Hopf Invariant One
An H-space is a space which admits a continuous product with unit. For the remaining notations, which are standard in homotopy-theory, we re- fer the reader to ...
[11]
[PDF] Group-like structures in general categories I multiplications and ...
I/X is an H-space then ~1 (X) is commutative; if X is a space o/the based homotopy type of a CW-eomplex and if X admits an R-structure then ~1 (X) is free. The ...<|control11|><|separator|>
[12]
[PDF] notes on the fundamental group - Harvard Mathematics Department
Let G be a group space and e ∈ G be the identity point. Then π1(G,e) is abelian. Proof. To show π1(G,e) is abelian, we will show that for any two loops γ,δ ...
[13]
H-space. The first is a commutator or generalized Samelson product ...
It is also possible to multiply maps f, g: XR – S to get f - g or fg: XR – S, where R and S are any spaces. This multiplication induces group structure in m (2R ...
[14]
[PDF] A primer of Hopf algebras - OSU Math
Borel [3]). In his 1941 paper [47], H. Hopf considered a more general situation. He called20 H-space any topological space X endowed with a continuous multi-.
[15]
[PDF] HOMOLOGY AND COHOMOLOGY OPERATIONS John Rognes 1 ...
For an H -space X , the algebra and coalgebra structures on H∗(X) induced by the Pontryagin product and the diagonal map are compatible, in the sense that the ...
[16]
[PDF] algebras can possibly be realized as the cohomology of finite H ...
This paper contains a calculation of H*Spin(n) as a Hopf algebra over tRe. Steenrod algebra A (where all homology and cohomology is to be taken with Z2.<|control11|><|separator|>
[17]
[PDF] jt3(S\ H. Hopf, W.K. Clifford, F. Klein
There are two parts to the paper: The first one is the definition of what now is called the Hopf invariant and the proof of its homotopy invariance. The second ...
[18]
https://www.jstor.org/stable/1969485
[19]
[PDF] Homologie Singuliere Des Espaces Fibres
3, November, 1951. Printed in U.S.A.. HOMOLOGIE SINGULIERE DES ESPACES FIBRES. Applications. PAR JEAN-PIERRE SERRE. Received May 4, 1951. INTRODUCTION. L'objet ...Missing: paper | Show results with:paper<|control11|><|separator|>
[20]
delooping in nLab
Sep 19, 2023 · For C = C = Top the (infinity,1)-category of topological spaces, a space is deloopable if it is an A-infinity-space and hence homotopy ...Idea · Definition · Characterization of deloopable... · Examples
[21]
[PDF] Infinite loop space theory - Department of Mathematics
The following diagram depicts some of the main concepts and the functors relating them. The word "coherent" in the previous paragraph refers, in part, to ...
[22]
Which homotopy 2-types are H-spaces? - MathOverflow
Oct 7, 2020 · Furthermore, if the necessary condition holds, vanishing of the Postnikov invariant β∈H3(π1(X),π2(X)) is a sufficient condition. Thus the ...Postnikov's algebraic reconstruction of cohomology from homotopy ...Spaces with same homotopy and homology groups that are not ...More results from mathoverflow.net
[23]
POSTNIKOV TOWERS WITH FIBERS GENERALIZED EILENBERG ...
→ πk(Xn) → πk(Xn−1) →···→ πk(X0). As is well known, the homotopy groups of X can be calculated from that of Xn in the tower by the Milnor exact sequence.
[24]
[PDF] Topological K-theory
Dec 15, 2016 · The idea of topological K-theory is that spaces can be distinguished by the vector bundles they support.
[25]
[PDF] Topological K-Theory - UChicago Math
2.1. H-spaces, division algebras, and tangent bundles of spheres. An H- space is a topological space equipped with a specific type of multiplicative ...
[26]
Unstable $ν_1$-Periodic Homotopy of Simply Connected, Finite $H
May 13, 2019 · Abstract page for arXiv paper 1905.05269: Unstable $ν_1$-Periodic Homotopy of Simply Connected, Finite $H$-Spaces, using Goodwillie Calculus.
[27]
[PDF] Goodwillie calculus - Michael Ching
Goodwillie calculus is a method for analyzing functors that arise in topology. One may think of this theory as a categorification of the classical ...