Flat Minima in Linear Estimation and an Extended Gauss Markov Theorem

Authors: Simon Segert

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Additionally, we analytically derive the generalization error in multiple random matrix ensembles, and compare with Ridge regression. Finally, we conduct an extensive simulation study, in which we show that the cross-validated Nuclear and Spectral regressors can outperform Ridge in several circumstances.
Researcher Affiliation Academia Simon Segert Princeton Neuroscience Institute Princeton University ssegert@princeton.edu
Pseudocode No The paper describes mathematical derivations and experimental procedures in text but does not include any explicit pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statements about releasing its source code or links to a code repository.
Open Datasets No The paper describes generating synthetic datasets from 'spherical Gaussian ensemble' and 'diagonal ensemble' (Section 2.2), and using 'Random Fourier features' (Section 3.2.2). These are not named, publicly available datasets with specific access information (links, DOIs, etc.).
Dataset Splits Yes For each model (Spectral, Ridge, Nuclear) and each individual dataset, we used 3-fold cross validation to select the best-performing on the training set.
Hardware Specification No The paper does not specify any hardware details (e.g., GPU models, CPU types, memory) used for running the experiments.
Software Dependencies No The paper mentions 'sklearn.kernel_approximation.RBFSampler' but does not provide specific version numbers for this or any other software dependencies.
Experiment Setup Yes In all cases, we used N = 100 training examples, and 5000 testing examples. ... The set of allowable values considered in the cross-validation was the same for all models, and consisted of 9 equally logarithmically spaced values spanning 10^-4 to 10^6.