Our approach introduces the first general Learning-to-Optimize (L2O) framework designed for Mixed-Integer Nonlinear Programming (MINLP). As illustrated above, the approach consists of two core ...
QIHD (Quantum-Inspired Hamiltonian Descent) is a quantum-inspired, GPU-enabled algorithmic framework for solving mixed-integer quadratic programming (MIQP) problems. MIQP is a classic optimization ...