Data fitting using Linear Programming function

Hello Everyone

I am new here, i had a question regarding linprog function. here is the description of the problem, below that is my attemtped code, please take a look at it, the problem i have is the constraint ap2 which should be 1.5, but putting it in a matrix form that would make the algorith work.

Now the reason i am here, is to seek assistance with this problem

suppose you want to fit the data set { tau(t); I(t); B(t)} available in the workspace, with the model

τau(t) = q1.(I^q2).(B^q3)

1. Reformulate the problem as a linear optimization one and introducethe constraints α2 < 0, α3 = 1.5 (Hint: transform the data using ln).
2. Write the corresponding Scilab algorithm using linpro

%Suppose we have the data {tau,I, B}

%initial values of the parameters, the parameters to be found
ap1=0.5;
ap2=-1;
ap3=1;

%the model equation Tau=ap1*(I^ap2)*(B^ap3);

% linearize by log

    log(tau)=Y;
    lap=log(ap1);
    Y= lap+ap2*log(I)+ap3*log(B)

    % the constraints are ap2<0; ap3=1.5;

    %the objective  c= Y-log(ap1)+ap2*log(I)+ap  3*log(B)

    c=[1;log(I);log(B)] % linprog solves minimization problems

    %alpha < 0
    ub= zeros(2);

    xsol=linprog(c,[],[],[],[],ub);

how can i otherwise approach the problem if my method isn't right.

Thank you a lot folks

Answer 1

[Sam]

Can anybody suggest how to approach this problem, please

I appreciate some suggestions