The Impact of Regularization on High-dimensional Logistic Regression
Authors: Fariborz Salehi, Ehsan Abbasi, Babak Hassibi
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The theory is validated by extensive numerical simulations across a range of parameter values and problem instances. |
| Researcher Affiliation | Academia | Fariborz Salehi, Ehsan Abbasi, and Babak Hassibi Department of Electrical Engineering California Institute of Technology Pasadena, CA, USA. |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement or link regarding the availability of its source code. |
| Open Datasets | No | For our analysis we assume that the regularizer f( ) is separable, i f(wi), and the data points are drawn independently from the Gaussian distribution, {xi}n i=1 i.i.d. N(0, 1 p Ip). Note that the assumptions considered in the analysis of the We further assume that the entries of β are drawn from a distribution Π. |
| Dataset Splits | No | The paper describes synthetic data generation for simulations, stating 'For the numerical simulations, the result is the average over 100 independent trials with p = 250 and κ = 1.' It does not specify train/validation/test dataset splits, as it does not use a fixed, pre-existing dataset that would typically be split for training and validation. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the numerical simulations. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers used for its numerical simulations. |
| Experiment Setup | No | The paper specifies parameters for numerical simulations such as 'average over 100 independent trials with p = 250 and κ = 1' and 'ϵ = 0.001'. However, these are parameters of the simulation environment and problem setup, not hyperparameter values or training configurations typically found in a machine learning experiment setup (e.g., learning rate, batch size, optimizer settings). |