I want to conduct a soil analysis using a different mix of let says Nutrition A, Nutrition B and Nutrition C.
Since I can put for each nutrition multiple values, I cannot try out all the possible combinations.
I am therefore looking for a simple optimization method, that provides a protocol for iteratively finding the optimal solution/or most optimal solution.
How would I conduct this analysis and is there some theory that can support this approach? I.e. concretely, how would one implement this in real life?
One guess that I had: Choose a best guess for each nutrition. Then change parameter A and evaluate the result. Continue until the parameter change does not improve the result. Then continue to adjust parameter B (with the best parameter A value), etc. etc.
However, this approach does not take into account interaction effects and I am worried that the result might be very different depending on which parameter I choose to adjust first.