Invariants of quivers under the action of classical groups (Q1014594)

\textit{M. Artin} conjectured in a paper [J. Algebra 11, 532--563 (1969; Zbl 0222.16007)] that the algebra of invariants of several \(m\times m\) matrices under the action of \(\text{GL}_m\) by simultaneous conjugation is generated by the traces of the monomials in the matrices. This was proved for a field of characteristic \(0\) by Sibirskii and Procesi independently. The appropriate version of the above conjecture in arbitrary characteristic is that the algebra of invariants is generated by the coefficients of the characteristic polynomial (i.e. by the traces of exterior powers) of the monomials in the matrices. This was first proved by \textit{S. Donkin} [Comment. Math. Helv. 69, No. 1, 137--141 (1994; Zbl 0816.16015)]. Le Bruyn and Procesi, and Schofield proved generalizations of the above result to the algebra of invariants of a quiver in characteristic \(0\). Here the space of tuples of matrices is replaced by the representation space of a quiver. The ``classical case'' of invariants of several matrices can be obtained by taking the one point quiver with \(n\) loops, where \(n\) is the number of matrices. \textit{M. Domokos} and \textit{A. N. Zubkov} [Transform. Groups 6, No. 1, 9--24 (2001; Zbl 0984.16023)] extended these results to arbitrary characteristic. Their method of proof was new and quite elementary. In the paper under review the author generalizes the notion of a quiver, to include other basic spaces whose invariants have been studied: spaces of forms and spaces of (anti)symmetric matrices. Furthermore, the action of products of classical groups is considered rather than just a product of general (or special) linear groups. The main result of the paper is Thm.~1 in Section 4. It gives generators for the algebra of invariants of a ``mixed quiver'' in arbitrary characteristic (the characteristic has to be \(\neq2\) if an orthogonal group is involved). To describe the basic invariants the notion of a \textit{tableau with substitutions} is needed, see Section~3.2. Sections 5-9 are dedicated to the proof of Thm.~1. Partial generalizations along the lines indicated above were already in the literature. Section~2.2 contains an overview of the known results.