Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classification
Authors: Evgenii Chzhen, Christophe Denis, Mohamed Hebiri, Luca Oneto, Massimiliano Pontil
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Finally, we present numerical experiments which indicate that our method is often superior or competitive with the state-of-the-art methods on benchmark datasets. and 5 Experimental results In this section, we present numerical experiments with the proposed method. |
| Researcher Affiliation | Academia | Evgenii Chzhen1,2, Christophe Denis1, Mohamed Hebiri1, Luca Oneto3, Massimiliano Pontil4,5 1Université Paris-Est, 2Université Paris-Sud, 3University of Pisa, 4Istituto Italiano di Tecnologia, 5University College London |
| Pseudocode | No | The paper describes methods using mathematical expressions and textual explanations but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | The source code we used to perform the experiments can be found at https://github.com/lucaoneto/NIPS2019_ Fairness. |
| Open Datasets | Yes | We consider the following datasets: Arrhythmia, COMPAS, Adult, German, and Drug2 and compare the following algorithms: Linear Support Vector Machines (Lin.SVM)... and For more information about these datasets please refer to [17]. |
| Dataset Splits | Yes | For Arrhythmia, COMPAS, German and Drug datasets we split the data in two parts (70% train and 30% test), this procedure is repeated 30 times... For the Adult dataset, we used the provided split of train and test sets. Unless otherwise stated, we employ two steps in the 10-fold CV procedure proposed in [17] to select the best hyperparameters with the training set5. |
| Hardware Specification | No | The paper does not provide any specific hardware details (e.g., GPU/CPU models, processor types, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper mentions software like 'Linear Support Vector Machines (Lin.SVM)', 'Logistic Regression (Lin.LR)', and 'Random Forests (RF)', and refers to Python code from other papers, but it does not provide specific version numbers for these software components or the programming language itself. |
| Experiment Setup | Yes | Unless otherwise stated, we employ two steps in the 10-fold CV procedure proposed in [17] to select the best hyperparameters with the training set5. ... The regularization parameter (for all method) and the RBF kernel with 30 values, equally spaced in logarithmic scale between 10 4 and 104. For RF the number of trees has been set to 1000 and the size of the subset of features optimized at each node has been search in {d, d 15/16 , d 7/8 , d 3/4 , d 1/2 , d 1/4 , d 1/8 , d 1/16 , 1} where d is the number of features in the dataset. |