Abstract: Disclosed herein are embodiments of a method performed by a hearing aid and a hearing aid. An acoustical model can be used to generate first values (hk) for the multiple frequency channels based on first parameters (G, A, l); and a statistical model can be configured to generate second values (rk) for the multiple frequency channels based on values of second parameters (s1, s2, . . . , sn) and a set of basis vectors (u1, u2, . . . , un). An optimizer can be configured to obtain an optimized set of parameter values, wherein the optimized set of parameter values can be obtained by minimizing a difference between the first sound pressure levels (robs) and a combination of the first values (rk) and the second values (hk).