Hi
I'm working on bacterial optimization for the degradation of heavy metals. I'v optimized some factors manually i-e 3-4 factors. But I'v to optimize some more factors. Do RSM suits me in optimizing 5 factors and If not, which other statistical/mathematical technique is better to use for the bacterial optimization?
Thanks
Farhan,
RSM works for 5 factors optimization, but you need to generate many experiments that capture the variate of your problem. It also depends on how many levels you want to determine for each factor.
Choosing the experimental design method is also important and influential.
The alternative methods for RSM is Ordinary Kriging that enables you to take the variation between response values into factors coefficient estimation.
Try to use computer generated optimal designs such as D-optimal design and IV-optimal design. They are much better than standard RSM. These designs are sophisticated and are flexible. Theses design will reduce the number of experiments.
If you want to use standard RSM, then try to apply screen design first to find few significant factors, then you can use RSM. This helps in reducing the number of experiments, while achieving good results.
Hi, there
Yes, you can. However, it is better to screen the five factors first using Blacket-Burman and later try using RSM.
Good luck
Karim salimi is right,
You should go for PB design first which will give you chance to select the most important medium constituents which influence the production of your interest. Then you can go for RSM by CCD or Box behnken. It is really useful for all types of production optimizations.
Follow these papers, you will get it.
http://link.springer.com/article/10.1134%2FS0026261715040037
http://www.sciencedirect.com/science/article/pii/S1878818115000729