Provably Auditing Ordinary Least Squares in Low Dimensions

Authors: Ankur Moitra, Dhruv Rohatgi

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Applying our algorithms to the Boston Housing dataset, we exhibit regression analyses where our estimator outperforms the greedy heuristic, and can successfully certify stability even in the regime where a constant fraction of the samples are dropped. 5 EXPERIMENTS
Researcher Affiliation Academia Ankur Moitra & Dhruv Rohatgi Massachusetts Institute of Technology {moitra, drohatgi}@mit.edu
Pseudocode Yes J FORMAL PSEUDOCODE FOR ALGORITHMS
Open Source Code No The paper does not contain any explicit statements about releasing source code or links to a code repository for the methodology.
Open Datasets Yes On the Boston Housing dataset (Harrison Jr & Rubinfeld, 1978), we regress house values against all pairs of features.
Dataset Splits No The paper states "On the entire dataset, we find a modest positive effect..." and "On the Boston Housing dataset... we regress house values against all pairs of features," implying the full dataset was used for analysis rather than explicit train/validation/test splits for model training.
Hardware Specification Yes All experiments were done in Python on a Microsoft Surface Laptop, using GUROBI (Gurobi Optimization, LLC, 2022) with an Academic License to solve the linear programs.
Software Dependencies No The paper mentions "Python" and "GUROBI (Gurobi Optimization, LLC, 2022)" but does not provide specific version numbers for these software dependencies.
Experiment Setup Yes Heterogeneous data experiment. For each dataset, we applied the net upper bound with 1000 trials, the LP lower bound with L = { 0.01, 0, 0.01} and m = 30, and the baseline lower bound with L = { 0.01, 0, 0.01} and m = 1000.