Coresets for Wasserstein Distributionally Robust Optimization Problems

Authors: Ruomin Huang, Jiawei Huang, Wenjie Liu, Hu Ding

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we implement our coreset approach and illustrate its effectiveness for several WDRO problems in the experiments.
Researcher Affiliation Academia 1School of Data Science 2School of Computer Science and Technology University of Science and Technology of China 3Department of Computer Science, City University of Hong Kong
Pseudocode Yes Algorithm 1 Dual ϵ-Coreset Construction
Open Source Code Yes The code is available at https://github.com/h305142/WDRO_coreset.
Open Datasets Yes We test the algorithms for the SVM and logistic regression problems on two real datasets: MNIST[28] and LETTER[8]. The dual coreset algorithm for the robust regression problem is evaluated on the real dataset APPLIANCES ENERGY[7].
Dataset Splits No The full settings of experiments are placed in the full version of this paper.
Hardware Specification Yes Our experiments were conducted on a server equipped with 2.4GHZ Intel CPUs and 256GB main memory.
Software Dependencies Yes We use the MOSEK [2] to solve the tractable reformulations of WDROs. [2] M. Ap S. MOSEK Optimizer API for Python 9.3.20, 2019.
Experiment Setup Yes We set c := s n to indicate the compression rate and fix the parameter γ = 7 for all the instances