Skip to main content

How bad are quadratic constraints?

Answered

Comments

2 comments

  • Official comment
    Simranjit Kaur
    • Gurobi Staff
    This post is more than two years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum, or try Gurobot, our chatbot interface offering instant, expert-level support.
  • Maliheh Aramon
    • Gurobi Staff

    Hi Florian, 

    Whether an MILP instance is easier to solve than an MIQP instance (quadratic objective with linear constraints) depends on several factors that are difficult to predict in advance. The best way to determine which formulation works better is to experiment with both. The consensus in the optimization community is that MILP instances are typically easier to solve than MIQP instances but there might be exceptions.

    Gurobi is highly capable of solving convex MIQP instances. If reformulating your problem as an MIQP reduces its size, this could result in a smaller and stronger presolved model. Therefore, if you are unsatisfied with the performance of your MILP formulation, we suggest considering a reformulation as an MIQP and testing it for comparison.

    Best regards,

    Maliheh

    0

Post is closed for comments.