Differentially Private Bayesian Linear Regression
Authors: Garrett Bernstein, Daniel R. Sheldon
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We design experiments to measure the calibration and utility of the private methods. |
| Researcher Affiliation | Academia | Garrett Bernstein University of Massachusetts Amherst gbernstein@cs.umass.edu Daniel Sheldon University of Massachusetts Amherst sheldon@cs.umass.edu |
| Pseudocode | Yes | Algorithm 1 Gibbs Sampler |
| Open Source Code | No | The paper mentions implementing methods using existing tools like PyMC3 but does not provide concrete access (link or explicit statement of availability) to the source code for the methodology described in this paper. |
| Open Datasets | Yes | We evaluate the predictive posteriors of the methods on a real world data set measuring the effect of drinking rate on cirrhosis rate.4 http://people.sc.fsu.edu/~jburkardt/datasets/regression/x20.txt |
| Dataset Splits | No | The paper states, 'There are 46 total points, which we randomly split into 36 training examples and 10 test points for each trial,' providing training and testing splits but no explicit separate validation split. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU/CPU models, processor types, or memory amounts used for running its experiments. |
| Software Dependencies | No | The paper mentions the use of 'Py MC3' but does not specify a version number or other software dependencies with their versions. |
| Experiment Setup | Yes | The noise-aware individual-based method (MCMC-Ind) is implemented using Py MC3 [Salvatier et al., 2016]; it runs with 500 burnin iterations and collects 2000 posterior samples. |