How do fair decisions fare in long-term qualification?
Authors: Xueru Zhang, Ruibo Tu, Yang Liu, Mingyan Liu, Hedvig Kjellstrom, Kun Zhang, Cheng Zhang
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our theoretical results and experiments on static real-world datasets with simulated dynamics show that our framework can be used to facilitate social science studies. |
| Researcher Affiliation | Collaboration | Xueru Zhang1, Ruibo Tu2, Yang Liu3 Mingyan Liu1 Hedvig Kjellström2 Kun Zhang4 Cheng Zhang5 1University of Michigan, {xueru,mingyan}@umich.edu 2KTH Royal Institute of Technology, {ruibo,hedvig}@kth.se 3University of California, Santa Cruz, yangliu@ucsc.edu 4Carnegie Mellon University, kunz1@cmu.edu 5Microsoft Research, Cheng.Zhang@microsoft.com |
| Pseudocode | No | The paper describes mathematical models and processes but does not include any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement or link indicating that source code for the described methodology is publicly available. |
| Open Datasets | Yes | We use the FICO score dataset [42] to study the long-term impact of fairness constraints Eq Opt and DP and other interventions on loan repayment rates in the Caucasian group GC and the African American group GAA. Our second set of experiments is conducted on a multivariate recidivism prediction dataset from Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) [3]. |
| Dataset Splits | No | The paper mentions using static datasets with simulated dynamics but does not provide specific training, validation, or test split percentages or sample counts for these simulations. It describes the process of simulating dynamics, but not data partitioning for reproducibility. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the experiments, such as CPU/GPU models or memory specifications. |
| Software Dependencies | No | The paper mentions training an "optimal classifier using a logistic regression model" but does not specify any software dependencies with version numbers (e.g., Python, PyTorch, scikit-learn, etc.). |
| Experiment Setup | No | The paper describes the general simulation process and model choices (e.g., beta distributions for feature distributions, logistic regression classifier), but it does not provide specific experimental setup details such as hyperparameter values (e.g., learning rates, batch sizes, number of epochs) that would be needed for reproduction. |