28  Concepts by Type

The map’s master diagram has one node per chapter, colored by which of five parts that chapter belongs to — Foundations, Objects, Structures, Theories, Applications. That’s why it can look sparse in a category like Objects: most objects (a vector, a matrix, a random variable) are discussed inside a structure or theory chapter rather than getting a page of their own, so they never became their own node. This page is the finer-grained companion: every concept in the glossary, classified individually, whether or not it has a chapter to itself.

A caveat worth stating rather than hiding: not everything below is cleanly an object, a structure, a theory, or an application — some entries are properties a structure can have (convex, bijective), maps between structures (homomorphism, linear map), or theorems about a theory (Bayes’ rule). Forcing those into one of the four boxes would be exactly the kind of false precision SPEC.md’s terminology discipline (§11) warns against, so the Notes column says so explicitly instead. The five section headings below match the map’s legend exactly; the notes carry the nuance the legend can’t.

28.1 Foundations

The substrate the object/structure/theory/application distinction is itself built on top of — logic, sets, functions, proof. Deliberately outside that four-way split, not a missing fifth box.

Concept Notes Chapter
Axiom a starting statement, not a claim needing proof Axioms and Definitions
Bijective / injective / surjective properties of a function Relations and Functions
Equivalence relation a relation with three specific properties Relations and Functions
Function a map — see What Is a Mathematical Structure? for why maps matter as much as objects Relations and Functions
Logic (proposition, proof) the inference rules everything else is checked against Logic and Proof
Mathematical induction a proof technique, introduced alongside its first real use Discrete Mathematics
Relation the general notion underlying equality, order, and functions alike Relations and Functions
Set the most primitive collection — almost everything else is built from it Sets
Theorem a statement derived from a theory’s axioms Axioms and Definitions

28.2 Objects

Specific instances — the level What Is a Mathematical Structure? uses “the matrix \(\begin{pmatrix}1&2\\0&1\end{pmatrix}\)” to illustrate. A general vector, a general matrix, and a general polynomial don’t get their own rows below — they’re the motivating examples inside Vector Spaces, not separate named concepts.

Concept Notes Chapter
Basis a specific spanning, linearly independent set of vectors Vector Spaces
Derivative a specific linear map (the best local approximation to \(f\)) Differentiation
Dimension a number attached to a vector space Vector Spaces
Eigenvalue / eigenvector a specific vector-and-scalar pair a linear map leaves undirected Linear Algebra
Embedding a specific learned vector Machine Learning
Entropy / KL divergence / mutual information numbers computed from a distribution Machine Learning
Expectation a number computed from a random variable Probability
Number systems (\(\mathbb{N},\mathbb{Z},\mathbb{Q},\mathbb{R},\mathbb{C}\)) the site’s one dedicated Objects chapter Number Systems
Random variable a specific function \(\Omega \to \mathbb{R}\) Probability
Span a specific set — all linear combinations of a given collection Vector Spaces
Zero divisor a specific ring element Rings and Fields

28.3 Structures

Sets equipped with operations or relations satisfying a fixed axiom list — What Is a Mathematical Structure?’s own subject, plus the maps between structures of the same kind (homomorphism, linear map, functor), grouped here rather than given a sixth category of their own.

Concept Notes Chapter
Abelian group a group with one extra property (commutativity) Groups
Affine map a map between affine spaces (linear map + translation) Affine Spaces
Affine space points and displacements, kept structurally apart Affine Spaces
Category the maximal abstraction of the homomorphism pattern Advanced Topics
Euclidean space affine space + inner product Euclidean Geometry
Field a ring where every nonzero element has a multiplicative inverse Rings and Fields
Functor a map between categories Advanced Topics
Group one operation, four axioms Groups
Hilbert space a complete inner-product space Advanced Topics
Homomorphism / isomorphism maps between structures of the same kind What Is a Mathematical Structure?
Inner product / norm angle/orthogonality, and the length it induces Inner Product Spaces
Linear map a map between vector spaces; a matrix once bases are fixed Linear Algebra
Mathematical structure the general pattern this whole section instantiates What Is a Mathematical Structure?
Metric space a set with a distance function Metric Spaces
Open set the basic building block a topology is made of Topological Spaces
Probability space outcomes, events, and a probability measure Probability
Ring two operations, related by distributivity Rings and Fields
Topological space a set with a designated collection of open sets Topological Spaces
Vector space addition and scaling Vector Spaces
Properties a structure can have: completeness, orthogonal, discrete/dense, convex, continuous not structures themselves — see each one’s own chapter

28.4 Theories

The body of results about a structure or family of problems — plus the named theorems and techniques that body of results actually produces.

Concept Notes Chapter
Bayes’ rule a theorem, not a structure Probability
Combinatorics counting subsets/functions of finite sets Discrete Mathematics
Continuous classically \(\varepsilon\)\(\delta\); topologically, preimages of opens Limits and Continuity
Convex / critical point properties/objects Optimization is built around Optimization
Differential Equations (ODE / PDE, dynamical system) equations relating a function to its own derivatives Differential Equations
Differentiation the theory built on the derivative Differentiation
Discrete / dense properties distinguishing \(\mathbb{N},\mathbb{Z}\) from \(\mathbb{Q},\mathbb{R}\) Discrete Mathematics
Euclidean Geometry affine + inner-product structure, assembled into geometry Euclidean Geometry
Functional analysis, measure theory, differential geometry, category theory preview-depth only, per SPEC.md §21 Advanced Topics
Fundamental theorem of calculus a theorem, not a structure Integration
Gradient descent an algorithm, not a structure Optimization
Independence a property/relation between random variables Probability
Integration (Riemann sum) the theory built on the integral Integration
Linear Algebra the theory of vector spaces and the linear maps between them Linear Algebra
Probability measure theory, wearing a specific interpretation Probability
Real analysis the rigorous foundation under Limits and Continuity Limits and Continuity

28.5 Applications

Domains that borrow a theory to solve a problem that isn’t, itself, mathematics.

Concept Notes Chapter
Backpropagation the chain rule, not separate mathematics Machine Learning
Machine Learning the site’s one written Applications chapter Machine Learning
Manifold hypothesis Euclidean Geometry’s manifold sketch, doing real work Machine Learning
Physics named throughout (Newton’s laws as ODEs, symmetry and invariance) but not yet its own chapter (planned)

28.6 Reading this table against the map

Where a row’s category surprised you, that’s usually the point: a derivative is an object (a specific linear map), not a theory, even though Differentiation — the theory about derivatives in general — is a whole chapter. That gap between “the theory” and “an instance the theory produces” is exactly what What Is a Mathematical Structure? spends a full chapter making explicit, and it’s the same distinction the map’s five-color legend is tracking at chapter granularity.