Clarifies relationships among MPEC constraint qualifications and fully characterizes MPEC-LICQ for the MPEC from bilevel hyperparameter optimization in L1-loss SVM classification.
Assume there exists nonzero column vector ρ defined by ρ = ρ⊤ 1 , ρ⊤ 2 ,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On Constraint Qualifications for MPECs with Applications to Bilevel Hyperparameter Optimization for Machine Learning
Clarifies relationships among MPEC constraint qualifications and fully characterizes MPEC-LICQ for the MPEC from bilevel hyperparameter optimization in L1-loss SVM classification.