A deep-unrolled PDQP network trained with an unsupervised KKT-residual loss predicts near-optimal QP solutions and accelerates the PDQP solver by up to 45%.
On implementing a primal-dual interior-point method for conic quadratic optimization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
An Efficient Unsupervised Framework for Convex Quadratic Programs via Deep Unrolling
A deep-unrolled PDQP network trained with an unsupervised KKT-residual loss predicts near-optimal QP solutions and accelerates the PDQP solver by up to 45%.