Using Function from Regression Learner into Optimization Toolbox

3 views (last 30 days)
Marc Youcef on 4 May 2017
Answered: Chaitral Date on 9 May 2017
Hello all,
Through the Regression Learner App, I have set a prediction function which is taking several parameters as predictors and outputting a numerical value. We will distinguish two sets of predictors used in this function: i1 (one predictor) and i2 (all the other predictors).
Our function looks like then : f(i1,i2)=Output.
What I know want to do is to minimize the following through Optimization toolbox: [f(i1,i2)-Target] where Target is a given constant. Minimization of this cost function should be done only by finding the optimal input value i1 which is giving the lowest value of f(i1,i2).
I do not know how to tell the Optimization toolbox to only play on input i1 of the f function to minimize [f(i1,i2)-Target]

Chaitral Date on 9 May 2017
Can you tell me how does your function look like? It is hard to tell without looking at the function.
Nevertheless, I do not find any of the functions in optimization toolbox which will only consider i1 from the f function to minimize [f(i1,i2)-Target]. You can use function "fminsearch" or "fminunc" to find min of [f(i1,i2)-Target] and can ignore i2 from the output.