What Are Lie Groups?

Lie groups are mathematical objects that are simultaneously groups and smooth manifolds, with the group multiplication and inversion operations required to be smooth maps. The definition joins two ideas that are usually studied separately: the algebraic structure of a group, which captures composition and symmetry, and the differential structure of a manifold, which permits calculus. Named for the Norwegian mathematician Sophus Lie, who introduced continuous transformation groups in the 1870s while looking for a Galois theory of differential equations, they provide the standard language for continuous symmetry in mathematics, physics, and engineering.

Familiar examples are all matrix groups. The rotation group SO(3) consists of three-by-three orthogonal matrices with determinant one and forms a three-dimensional compact manifold. The special Euclidean group SE(3) combines rotation with translation and describes rigid body motion in space. The unitary groups U(n) and SU(n) describe phase and internal symmetries in quantum mechanics, and each of these is realized as a subgroup of a general linear group GL(n) of invertible matrices.

Manifold and Group Structure

Because a Lie group is a manifold, every element has a neighborhood that looks like an open set in Euclidean space, which means that group elements can be differentiated, integrated over, and optimized. Because it is also a group, that local structure is the same everywhere: left multiplication by a fixed element is a diffeomorphism carrying any neighborhood of the identity onto a neighborhood of any other point. This homogeneity is what makes Lie groups tractable, since a global statement about the group can often be reduced to a statement about the identity element. Compactness is the other structural property that matters in practice, separating groups such as SO(3) and SU(2), which are compact and admit an invariant finite measure, from noncompact groups such as the Lorentz group O(3,1).

Lie Algebras and the Exponential Map

The tangent space at the identity, equipped with a bracket operation derived from the group commutator, is the Lie algebra of the group. It is a linear vector space, and it captures nearly all of the group's local structure. The exponential map sends elements of the algebra to elements of the group, so a curved nonlinear group can be handled with linear algebra by working in the algebra and mapping back. This correspondence is the working tool in applied fields, and a practical treatment of Lie theory for state estimation in robotics sets out the exponential map, the logarithm, adjoint operators, and the Jacobians needed to propagate uncertainty on SO(3) and SE(3). More formal developments, such as a set of graduate lecture notes on Lie groups, build the same machinery from the differential-geometric side and extend it to Lie group actions on manifolds.

Representations, Classification, and Symmetry Analysis

A representation of a Lie group is a homomorphism into the group of invertible linear operators on a vector space, which lets abstract symmetry act on concrete data. Representation theory of compact groups is well developed: every finite-dimensional representation decomposes into irreducible pieces, and the irreducibles of SU(2) are indexed by a label taking integer and half-integer values, which is the mathematical origin of quantum mechanical spin. The simple Lie algebras over the complex numbers were classified by Killing and Cartan into four infinite families and five exceptional cases, a result that underpins the group structure of the Standard Model. Lie's original motivation also remains active: symmetry group analysis of differential equations uses continuous symmetries of an equation to reduce its order, generate new solutions from known ones, and identify conserved quantities through Noether's theorem.

Applications

Lie groups have applications in a wide range of disciplines, including:

• Rigid body kinematics, robot pose estimation, and SLAM
• Attitude determination and control of spacecraft
• Gauge field theory and particle physics
• Geometric control and optimization on manifolds
• Computer vision, including camera pose and shape analysis
• Signal processing on spheres and rotation groups