Domain concepts for quadratic programming. No API or interface details here. QP support in cuOpt is currently in beta.

What is QP

Objective: Quadratic in the variables (e.g. x², x·y terms). Example: portfolio variance xᵀQx.
Constraints: Linear only. cuOpt does not support quadratic constraints.

Important domain rule: minimize only

QP objectives must be minimization. To maximize a quadratic expression, negate it and minimize; then negate the optimal value.

Ask these if not already clear:

Objective — Does it have squared or cross terms (x², x·y)? If purely linear, use LP/MILP instead.
Minimize or maximize? — If maximize, user must negate objective and minimize.
Convexity — For minimization, the quadratic form (matrix Q) should be positive semi-definite for well-posed problems.

Domain concepts for quadratic programming. No API or interface details here. QP support in cuOpt is currently in beta.

Objective: Quadratic in the variables (e.g. x², x·y terms). Example: portfolio variance xᵀQx.
Constraints: Linear only. cuOpt does not support quadratic constraints.

QP objectives must be minimization. To maximize a quadratic expression, negate it and minimize; then negate the optimal value.

Ask these if not already clear:

Objective — Does it have squared or cross terms (x², x·y)? If purely linear, use LP/MILP instead.
Minimize or maximize? — If maximize, user must negate objective and minimize.
Convexity — For minimization, the quadratic form (matrix Q) should be positive semi-definite for well-posed problems.