A Modified Orthant-Wise Limited Memory Quasi-Newton Method with Convergence Analysis

Authors: Pinghua Gong, Jieping Ye

ICML 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also provide empirical studies to show that m OWLQN works well and is as efficient as OWL-QN. In this section, we validate the convergence analysis by applying the m OWL-QN algorithm to solve the following ℓ1regularized logistic regression problem: ... Experiments are conducted on four large scale data sets which are summarized in Table 1. ... We report the objective function value vs. CPU time plots in Figure 1.
Researcher Affiliation Academia Pinghua Gong GONGP@UMICH.EDU University of Michigan, Ann Arbor, MI 48109 Jieping Ye JPYE@UMICH.EDU University of Michigan, Ann Arbor, MI 48109
Pseudocode Yes Algorithm 1 OWL-QN: Orthant-Wise Limited memory Quasi-Newton; Algorithm 2 m OWL-QN: modified Orthant-Wise Limited memory Quasi-Newton
Open Source Code No The paper does not include an unambiguous statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes These data sets are high-dimensional and sparse and can be downloaded from http://www.csie.ntu.edu.tw/ cjlin /libsvmtools/datasets/binary.html.
Dataset Splits No The paper uses large-scale datasets but does not specify how these datasets were split into training, validation, or test sets, nor does it mention cross-validation. The focus is on the optimization algorithm itself.
Hardware Specification Yes Both algorithms are implemented in Matlab and executed on an Intel(R) Core(TM)2 i7-3770 CPU (@3.4GHz) with 32GB memory.
Software Dependencies No The paper states that algorithms are 'implemented in Matlab' but does not specify a version number for Matlab or any other specific software libraries or dependencies used, which would be necessary for full reproducibility.
Experiment Setup Yes We terminate both algorithms if the relative change of two consecutive objective function values is less than 10 5 or the number of iterations exceeds 500. We set γ = 10 2, β = 0.2, α0 = 1, ϵ = 10 12 and the number of unrolling steps (in L-BFGS) as m = 10.