Good invariants for bad ideals (Q984425): Difference between revisions

\textit{M.~P.~Murthy} [Ann. Math. (2) 140, No. 2, 405--434 (1994; Zbl 0839.13007)] defined a suitable Segre class of a finitely generated module \(M\) over a smooth affine domain \(A\) over an algebraically closed field, \(s_0(M)\in CH_0(A)\). This class was shown to be the obstruction for the module \(M\) to satisfy a suitable Eisenbud-Evans estimate as its number of generators. If one is working over arbitrary rings (not necessarily affine domains over algebraically closed fields) it was well known that the above class is insufficient to give the right estimates. The philosophy has been, in this case, one should define the Segre class in a suitable Euler class group (not the Chow group). This plan the authors carry out for suitable ideals and prove similar theorems as in the case dealt with by Murthy. This allows the authors to prove certain `addition' and `subtraction' theorems, which can be loosely termed as considering `residual intersections', a method as old as algebraic geometry.