Certifying Fairness of Probabilistic Circuits
Authors: Nikil Roashan Selvam, Guy Van den Broeck, YooJung Choi
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate our discrimination pattern mining algorithms on three datasets: COMPAS which is used for recidivism prediction and the Adult (Dua and Graff 2017) and Income (Ding et al. 2021) datasets for predicting income levels. |
| Researcher Affiliation | Academia | 1 Computer Science Department, University of California, Los Angeles 2 School of Computing and Augmented Intelligence, Arizona State University |
| Pseudocode | Yes | Algorithm 1 outlines the pseudocode of our search algorithm. (...) Algorithm 2 SAMPLE-DISC-PATTERNS(C, Z) |
| Open Source Code | Yes | Link to pre-processed data, trained models, and code: https://github.com/UCLA-Star AI/PC-DiscriminationPatterns. |
| Open Datasets | Yes | We evaluate our discrimination pattern mining algorithms on three datasets: COMPAS which is used for recidivism prediction and the Adult (Dua and Graff 2017) and Income (Ding et al. 2021) datasets for predicting income levels. (...) Dua, D.; and Graff, C. 2017. UCI Machine Learning Repository. http://archive.ics.uci.edu/ml. Accessed: 2022-01-06. (...) Ding, F.; Hardt, M.; Miller, J.; and Schmidt, L. 2021. Retiring adult: New datasets for fair machine learning. Advances in Neural Information Processing Systems, 34. |
| Dataset Splits | No | The paper mentions using three benchmark datasets but does not specify the train/validation/test split percentages or methodology used for its experiments. |
| Hardware Specification | Yes | All experiments were run on an Intel(R) Xeon(R) CPU E5-2640 (2.40GHz). |
| Software Dependencies | No | The paper mentions using the STRUDEL algorithm but does not provide specific version numbers for software dependencies or libraries used. |
| Experiment Setup | No | The paper describes the algorithms and evaluation on datasets but does not provide specific hyperparameter values, training schedules, or other detailed system-level experimental setup configurations. |