Compact Operators - Core Definitions
Compact operators generalize finite-rank operators and matrices to infinite dimensions, providing a class of operators with particularly nice spectral properties.
Let and be Banach spaces. A bounded linear operator is compact if for every bounded sequence in , the sequence has a convergent subsequence in .
Equivalently, is compact if it maps the closed unit ball of to a relatively compact subset of .
Compact operators can be thought of as "almost finite-dimensional" operators. They map bounded sets to precompact sets, exhibiting behavior closer to finite-dimensional linear algebra than general bounded operators.
-
Finite Rank: Any operator with finite-dimensional range is compact
-
Integral Operators: The operator on with continuous kernel is compact
-
Diagonal Operators: On , the operator is compact if and only if
-
Inclusion Maps: The inclusion is compact (Rellich-Kondrachov theorem)
-
Hilbert-Schmidt Operators: On , operators with kernels are compact
Let denote the space of compact operators from to . Then:
- is a closed subspace of
- If is compact and is bounded, then and are compact
- is compact if and only if is compact
- Limits of finite-rank operators are compact
- On infinite-dimensional spaces, the identity operator is never compact
Property 5: If were compact on an infinite-dimensional space, the unit ball would be precompact, hence compact. But by the Riesz Lemma, the unit ball in an infinite-dimensional normed space is never compact.
Compact operators form an ideal in the algebra of bounded operators: products of compact with bounded operators are compact. This algebraic structure is crucial for perturbation theory and operator equations.
The study of compact operators provides a bridge between finite-dimensional linear algebra and infinite-dimensional functional analysis, with applications to integral equations, spectral theory, and PDEs.