Problem related to the prediction equations

[GAURAV VERMA Res. Scholar, Dept. of Civil Engg., IIT (BHU)]

Dear All,

I am looking for a solution related to a problem faced in the prediction equation obtained through the chromosome solution in MEP software. Among a total of 10 runs programmed for a particular problem, I have selected one best run, based on the error value obtained for training, testing and validation dataset. For the selected run, the output value shown in the software for each of the observations was noted and the respective chromosome solution (code) was saved. Subsequently, the code was solved and the predicting equation was generated. Unfortunately, it has been found that the output value obtained through the generated equation doesn't match with the previously noted output values. This particular problem has been faced many a time in the software. May anyone suggest, why this is happening so?

Regards

Gaurav Verma

[mihai....@gmail.com]

Hi,

Please save the project and send it to me as an attachment (directly to my address: mihai....@gmail.com ).

waiting,

Mihai

[GAURAV VERMA Res. Scholar, Dept. of Civil Engg., IIT (BHU)]

Dear all,

It was not a problem from the MEPX software side, but the mistake that I was doing in solving the code achieved through MEPX software. I am really thankful to Dr. Mihai Oltean who has solved the problem with remarkable interests. Moreover, I am greatly thankful to you for providing some significant information about the MEPX software.

Thank you

Regards

Gaurav Verma