JOM-e Logo The following is a supplement to the article "Extrusion Die Design: A New Methodology for Experimental Design as a Precursor to Neural Networks" by Bhavin V. Mehta et al., which is as part of JOM-e, 51 (9) (1999).

Table VIII. Error Indicators Results for the Training and Testing Sets
For the Training Set Before the D.O.E. After the D.O.E.
Average Error -0.0409 -0.0149
Average Absolute Error 1.8647 2.2138
Average Percentage Error 1.3655 1.4879
Mean Square Error 0.8174 0.8552
Root Mean Square Error 10.3798 12.1502
Maximum Absolute Error -6.3547 -5.4214
For the Testing Set Before the D.O.E. After the D.O.E.
Average Error -0.0785 -0.0391
Average Absolute Error 1.6939 2.0887
Average Percentage Error 1.3015 1.4452
Mean Square Error 0.7942 0.8507
Root Mean Square Error 10.3798 12.1465
Maximum Absolute Error -8.7345 -8.0900

Copyright held by The Minerals, Metals & Materials Society, 1999

Direct questions about this or any other JOM page to jom@tms.org.

Search TMS Document Center Mehta's Main Text Other Hypertext Articles JOM TMS OnLine