TUDelft modeling/26 October 2008

=October 26st=

I have tried many different inhibition functions. As Rad wrote before the Marco suggestion was also examined and it was better than the one with the plus but still not satisfactory and with the problem of the root. After trying different functions finally I reached to a combination of polynomials and error function in the rational format which could estimate the lab results with SSE of less than 1. I don't know if there should be any kind of restriction on the inhibition function (I know some general ones like as the limit of f(x) and f'(x) go to zero when x goes to infinity). I will come tomorrow to see if it is OK to use that as the inhibition function.

Farzad