Statistical Properties of Robust Satisficing
Authors: Zhiyi Li, Yunbei Xu, Ruohan Zhan
ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our numerical experiments show that the RS model consistently outperforms the baseline empirical risk minimization in small-sample regimes and under distribution shifts. |
| Researcher Affiliation | Academia | 1School of Mathematical Sciences, Peking University, Beijing, China 2Yunbei Xu, Laboratory for Information and Decision Systems, Massachusetts Institute of Technology, Cambridge, MA, USA 3Department of Industrial Engineering and Decision Analytics, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong SAR 4HKUST Shenzhen-Hong Kong Collaborative Innovation Research Institute, Futian, Shenzhen, China. |
| Pseudocode | No | The paper does not include any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any concrete access to source code, such as a repository link or an explicit statement about code release. |
| Open Datasets | No | The paper describes generating its own data: 'the feature variable u is drawn from a normal distribution: u N [0.5, 0.5, ..., 0.5]T , 0.5Imu ; and the label variable y is generated via a linear model: y = u x + e'. No public dataset access information (link, DOI, or specific citation to a pre-existing dataset) is provided. |
| Dataset Splits | No | The paper mentions 'training data' but does not provide specific dataset split information (e.g., percentages, sample counts, or methodology for creating train/validation/test sets) needed to reproduce data partitioning. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory, or cloud instance types) used for running its experiments. |
| Software Dependencies | No | The paper does not specify any software dependencies with version numbers (e.g., library names like PyTorch 1.9, or solver names like CPLEX 12.4). |
| Experiment Setup | Yes | For this purpose, we consider a relatively highdimensional setting with the dimension mu = 10 and the true model parameter x = x = [2.0, 1.0, ..., 2.0, 1.0]T .We vary the tolerance rate ϵ in the RS model and the radius r in the DRO model. We set the model parameter in the sampling distribution to x = [2.0, 1.0]T , as in Section 5.2; and set the target environment to be x = [1.80, 0.90]T . |