estimated variance on parameter fit
- To: mathgroup at smc.vnet.net
- Subject: [mg127277] estimated variance on parameter fit
- From: Josh <jhcooperman at gmail.com>
- Date: Fri, 13 Jul 2012 02:54:42 -0400 (EDT)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- Delivered-to: mathgroup-newout@smc.vnet.net
- Delivered-to: mathgroup-newsend@smc.vnet.net
I am trying to determine the optimal fit of a square matrix A of data to a square matrix B of fit parameters b11, b12, . . .. Using NMinimize on the square of the trace of the difference of these two matrices works quite well. That is, NMinimize[Tr[MatrixPower[A-B,2]],{b11, b12, . . . }]. I would like to input the measurement errors on the data, perhaps as weights for the fit, and have Mathematica output the estimated errors on its determinations of the parameters b11, b12, . . .. Does anyone know how to accomplish this?