Bilevel learning of the Group Lasso structure

Authors: Jordan Frecon, Saverio Salzo, Massimiliano Pontil

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The performance of the proposed approach are quantitatively assessed on synthetic data in Section 4, and shown to favorably compare against standard approaches. In addition, an application to real data in the context of gene expression analysis is provided with the goal of discovering functional groups. In this section, we first devise synthetic experiments to illustrate and assess the performance of the proposed method. Then, we tackle a real-data experiment in the context of gene expression analysis.
Researcher Affiliation Academia 1 Computational Statistics and Machine Learning, Istituto Italiano di Tecnologia (Italy) 2 Department of Computer Science, University College London (UK)
Pseudocode Yes Algorithm 1 Dual forward-backward with Bregman distances: FBB-GLasso(y, X, λ, θ)
Open Source Code No A MATLAB R toolbox is available upon request to the authors.
Open Datasets Yes In this section, we lead a preliminary experiment on gene expression data collected from https://www.ensembl.org/ using Bio Mart.
Dataset Splits Yes The data set is split into training, validation and test sets of 20 genes each.
Hardware Specification No The paper does not provide any specific hardware details such as GPU models, CPU types, or cloud computing specifications used for experiments.
Software Dependencies No The paper mentions a 'MATLAB R toolbox' but does not specify any version numbers for MATLAB or any other software dependencies.
Experiment Setup Yes We consider the convex relaxation pointed in Remark 2.1, set (Q = 500, ϵ = 10 3, γ = 0.1, K = 2000) and denote the proposed solution as θBi GL.