A Primal-Dual-Assisted Penalty Approach to Bilevel Optimization with Coupled Constraints

Authors: Liuyuan Jiang, Quan Xiao, Victor Tenorio, Fernando Real-Rojas, Antonio G. Marques, Tianyi Chen

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We establish rigorous convergence theory for the proposed algorithm and demonstrate its effectiveness on two well-known real-world applications hyperparameter selection in support vector machine (SVM) and infrastructure planning in transportation networks using the real data from the city of Seville.
Researcher Affiliation Academia Rensselaer Polytechnic Institute, Troy, NY, United States King Juan Carlos University, Madrid, Spain
Pseudocode Yes Algorithm 1 Meta algorithm: BLOCC
Open Source Code Yes The code is available at https://github.com/Liuyuan999/Penalty Based Lagrangian Bilevel.
Open Datasets Yes We test the performance of our algorithm BLOCC with γ = 12 when training a linear SVM model on the diabetes [20] and fourclass datasets [29].
Dataset Splits Yes The experiments are executed for 50 different random train-validation-test splits, with the bold line representing the mean, and the shaded regions being the standard deviation.
Hardware Specification No The paper does not provide specific hardware details such as GPU/CPU models, memory, or specific computing cluster configurations used for running the experiments.
Software Dependencies No The paper mentions software by name (e.g., BLOCC, LV-HBA, GAM) and sets hyperparameters for them, but does not specify version numbers for any software components, libraries, or programming languages.
Experiment Setup Yes We use γ = 12 and η = 0.01 for running our BLOCC and we apply α = 0.01, γ1 = 0.1, γ2 = 0.1, η = 0.001 for LV-HBA and α = 0.05, ϵ = 0.005 for GAM respectively.