IndiJones
Full Member level 2
neural network specialist
I want to minimize the error between the measured and output values. I have 10 sets of data (measured). Each set contains three values Xm, Ym, Zm. So the measured values form a 10X3 matrix. The corresponding output values for each set of data are S1, S2 and S3 (also a 10X3 matrix). Multiplying the transpose of the S1S2S3 matrix by a 3X3 matrix gives the final output values Xo, Yo and Zo for each set of data. I want to find this 3X3 matrix that would give values of Xo, Yo and Zo closest to those of Xm, Ym and Zm, respectively.
At present I'm using multiple regression analysis. I was wondering if it's possible to a better job using Neural Nets.
Thanks in advance!
I want to minimize the error between the measured and output values. I have 10 sets of data (measured). Each set contains three values Xm, Ym, Zm. So the measured values form a 10X3 matrix. The corresponding output values for each set of data are S1, S2 and S3 (also a 10X3 matrix). Multiplying the transpose of the S1S2S3 matrix by a 3X3 matrix gives the final output values Xo, Yo and Zo for each set of data. I want to find this 3X3 matrix that would give values of Xo, Yo and Zo closest to those of Xm, Ym and Zm, respectively.
At present I'm using multiple regression analysis. I was wondering if it's possible to a better job using Neural Nets.
Thanks in advance!