Skip to main content

Estimate a Coefficient from Objective Function so Optimization Result Fits Data (Tuning Model)



1 comment

  • Riley Clement
    Gurobi Staff Gurobi Staff

    Hi Jack,

    I don't have any advice regarding estimating theta, hopefully someone else will, but I wanted to quickly point out that your original problem can be formulated without non-linear constraints:

    \max && \sum_{i=1}^n w_iX_i + 0.7Y_i + (X_i + Y_i)\theta\\
    s.t. &&\\
    && \sum_{i=1}^n (0.7 - w_i)X_i \geq 0\\
    && \sum_{i=1}^n (w_i - 0.7)Y_i \geq 0\\
    && X_i + Y_i \leq 1, \quad \forall i\\

    There aren't any tricks here, just rearranging the constraints.

    - Riley



Please sign in to leave a comment.