On Prior Distributions and Approximate Inference for Structured Variables

Authors: Oluwasanmi O Koyejo, Rajiv Khanna, Joydeep Ghosh, Russell Poldrack

NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on simulated data and high dimensional neuroimaging data show the effectiveness of our approach in terms of support recovery and predictive accuracy.
Researcher Affiliation Academia Oluwasanmi Koyejo Psychology Dept., Stanford sanmi@stanford.edu Rajiv Khanna ECE Dept., UT Austin rajivak@utexas.edu Joydeep Ghosh ECE Dept., UT Austin ghosh@ece.utexas.edu Russell A. Poldrack Psychology Dept., Stanford poldrack@stanford.edu
Pseudocode No The paper describes algorithms (e.g., greedy forward selection) but does not present them in a structured pseudocode or algorithm block format.
Open Source Code No The paper mentions using publicly available implementations for baselines (scikit-learn, Spike and Slab) but does not provide access to its own source code.
Open Datasets Yes FMRI data were collected from 126 participants while the subjects performed a stop-signal task [24]. ... f MRI data was provided by the Consortium for Neuropsychiatric Phenomics (NIH Roadmap for Medical Research grants UL1-DE019580, RL1MH083269, RL1DA024853, PL1MH083271).
Dataset Splits Yes For each of these models, the hyperparameter was selected in an inner 5-fold cross validation loop.
Hardware Specification No The paper does not explicitly specify the hardware (e.g., CPU, GPU models) used for running the experiments.
Software Dependencies No The paper mentions software like 'scikit-learn python package [20]' and 'publicly available implementation of Spike and Slab [21]' but does not provide specific version numbers for these dependencies.
Experiment Setup Yes The noise variance hyperparameter for Ridge and ARD were selected from the set 10{ 4, 3,...,4}. Lasso was evaluated using the default scikit-learn implementation where the hyperparameter is selected from 100 logarithmically spaced values based on the maximum correlation between the features and the response.