On Coxeter-Dynkin diagrams of hypersurface singularities (Q1356256)

Let \(f:(\mathbb{C}^n,0)\to(\mathbb{C},0)\) define an isolated hypersurface singularity \((n\equiv 3\bmod 4)\). Let \((\delta_1,\dots,\delta_\nu)\) be a distinguished basis of vanishing cycles of \(f\). The Coxeter-Dynkin diagram \(D\) of the singularity defined by \(f\) with respect to \((\delta_1,\dots,\delta_\mu)\) is the graph which has as vertices the cycles \(\delta_i\) and the vertices are joined by an edge of weight \((\delta_i,\delta_j)\), the intersection number of the cycles \(\delta_i\) and \(\delta_j\), if \((\delta_i,\delta_j)\neq 0\). Let \(\rho(D)\) be the smallest number of edges of \(D\) which have to be removed in order to delete all proper monotone cycles of \(D\), and let \(\eta(D)\) be the number of edges of negative weight of \(D\). There are two conjectures of Il'yuta: 1. \(m(f)=\min\{\rho(D)|\) \(D\) a Coxeter-Dynkin diagram of \(f\}\); 2. \(m(f)=\min\{\eta(D)|\) \(D\) a Coxeter-Dynkin diagram of \(f\}\); here \(m(f)\) denotes the modality of \(f\). Il'yuta proved this for the 0- and unimodal singularities. It is proved that both conjectures are true in a stronger form for 0- and unimodal singularities.