Algebras of linear transformations (Q1589830): Difference between revisions

The book under review is about associative algebras and their representations as algebras of linear transformations. Almost exclusively, the algebras under consideration are finite-dimensional, hence the representations are also finite-dimensional. The field of scalars is often either the complex field or the real field. The material is built around several key ideas and results, among them: Frobenius theorem on finite-dimensional real division algebras, invariant subspaces and Burnside's theorem, structure theorems for simple and semisimple algebras, operator algebras and Gelfand-Naimark theorem (the second name is spelled Neumark in the book), tensor products. The subject matter is classical, and is found in many books. However, the approach and selection of material is novel. In contrast with many texts that offer detail discussions of finite-dimensional algebras in a ``top-down'' manner, namely, only after a substantial theory of rings and modules, the approach offered in the book under review is ``bottom-up'', namely, presented as an extension and further development of linear algebra. Another novelty of exposition is a thorough discussion of finite-dimensional von Neumann and involutive algebras in the general context of associative algebras. The book consists of 6 chapters. Chapter 1: Linear algebra. Several linear algebra concepts and results are reviewed, culminating with the spectral theorem for normal linear transformations in a finite-dimensional inner product space. Chapter 2: Algebras. Representations, simple algebras, division algebras, and a complete proof of Frobenius theorem. Chapter 4: Invariant subspaces. Burnside's theorem, and results on triangularization of commutative subalgebras and of semigroups of nilpotents, are the highlights of this chapter. Chapter 4: Semisimple algebras. The chapter begins with discussion on nilpotent algebras (key result: only nilpotent algebras do not have nonzero idempotents) and the radical, and proceeds to the classification of semisimple algebras with complete proofs. Chapter 5: Operator algebras. The focus here is on von Neumann algebras, defined as unital subalgebras of the algebra of linear transformations on a finite-dimensional complex inner product space that are closed under the operation of taking adjoints. A full proof is given of Gelfand-Naimark theorem that asserts that every finite-dimensional complex algebra with positive involution admits a faithful *-representation as a von Neumann algebra. A real analogue of this result is proved as well. A Wedderburn-type theorem is proved for finite-dimensional real operator algebras (i.e., involutive unital real algebras such that \(a^*a= 0\) implies \(a=0\) and the element \(1+ a^*a\) is invertible for every \(a\) in the algebra): Every such algebra is *-isomorphic to a direct sum of full matrix algebras over the reals, the complex numbers, or the quaternions. The last chapter is devoted to the tensor product and its base properties. Simplicity and semisimplicity properties of real tensor products of full matrix algebras are discussed. The exposition is appropriate for lower level of graduate school, although formally the prerequisites are minimal, namely, basic undergraduate courses in linear algebra and algebra. Thus, upper level undergraduates may find the book accessible as well. All theorems, lemmas, etc. (with very few exceptions) are provided with complete proofs. The text is frequently interspersed with well-selected examples to underscore significance and limitations of results under discussion. Group algebras over finite groups provide a steady supply of interesting examples and exercises. A few relevant major theorems, such as Wedderburn's theorem on finite division rings, whose proofs the author deemed to be out of the scope of the book, are discussed with a reference to the proofs. The book contains 125 exercises, mostly of theoretical nature, that significantly augment the material in the text. The book is written in a lively style, with meticulous attention to detail, and is well-suited as a basis for a lower level graduate course, or an upper level undergraduate course for well-motivated students, as well as for self-study. It can serve, together with a basic course in functional analysis, as a preparation for study of infinite-dimensional operator algebras. The bibliography is kept to a bare minimum of 30 items.