An Improved Analysis of Alternating Minimization for Structured Multi-Response Regression

Authors: Sheng Chen, Arindam Banerjee

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results support our theoretical developments.
Researcher Affiliation Collaboration Sheng Chen The Voleon Group chen2832@umn.edu Arindam Banerjee Dept. of Computer Science & Engineering University of Minnesota, Twin Cities banerjee@cs.umn.edu
Pseudocode Yes Algorithm 1 Alternating minimization for multi-response regression
Open Source Code No The paper does not provide an explicit statement or link for open-source code availability.
Open Datasets No The paper uses synthetically generated data: 'Entries of X of η are generated by i.i.d. standard Gaussian, and θ = [1, . . . , 1 | {z } 10 , 1, . . . , 1 | {z } 10 , 0, . . . , 0 | {z } 980 ]T . Σ is given as a block diagonal matrix with Σ = h 1 a a 1 i replicated along the diagonal.' No access information is provided.
Dataset Splits No The paper mentions 'sample size n' but does not specify training, validation, and test dataset splits with percentages or counts for their experiments.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models) used for running the experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers.
Experiment Setup Yes Specifically we focus on the sparsity structure of θ , and consider L0-cardinality as complexity function f. Throughout the experiment, we fix problem dimension p = 1000, sparsity level of θ s = 20, and number of iterations T = 10. Entries of X of η are generated by i.i.d. standard Gaussian, and θ = [1, . . . , 1 | {z } 10 , 1, . . . , 1 | {z } 10 , 0, . . . , 0 | {z } 980 ]T . Σ is given as a block diagonal matrix with Σ = h 1 a a 1 i replicated along the diagonal. All the plots are obtained based on the average over 100 random trials.