High Dimensional Structured Superposition Models

Authors: Qilong Gu, Arindam Banerjee

NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we confirm the theoretical results in this paper with some simple experiments. We show our experimental results under different settings. In our experiments we focus on MCA when k = 2. The design matrix X are generated from Gaussian distribution such that every entry of X subjects to N(0, 1). The noise ω is generated from Gaussian distribution such that every entry of ω subjects to N(0, 1). We implement our algorithm 1 in MATLAB. We use synthetic data in all our experiments, and let the true signal θ1 = (1, . . . , 1 | {z } s1 , 0 . . . , 0), Qθ2 = (1, . . . , 1 | {z } s2 , 0 . . . , 0). We generate our data in different ways for our three experiments. Figure 2: (a) Effect of parameter ρ on estimation error when noise ω = 0. We choose the parameter ρ to be 0, 1/ 2, and a random sample. (b) Effect of dimension p on fraction of successful recovery in noiseless case. Dimension p varies in {20, 40, 50, 150}
Researcher Affiliation Academia Qilong Gu Dept of Computer Science & Engineering University of Minnesota, Twin Cities guxxx396@cs.umn.edu Arindam Banerjee Dept of Computer Science & Engineering University of Minnesota, Twin Cities banerjee@cs.umn.edu
Pseudocode No The paper presents the estimator as a mathematical optimization problem (2) but does not provide it in a pseudocode or algorithm block format.
Open Source Code No The paper does not include any statements or links indicating that open-source code for the methodology is provided.
Open Datasets No We use synthetic data in all our experiments, and let the true signal θ1 = (1, . . . , 1 | {z } s1 , 0 . . . , 0), Qθ2 = (1, . . . , 1 | {z } s2 , 0 . . . , 0).
Dataset Splits No The paper mentions using synthetic data and varying sample sizes, but it does not specify any train/validation/test dataset splits. For example, it says 'The number of samples n varied between 1 and 1000.' and 'for each n, we generate 100 pairs of X and w.'
Hardware Specification No The paper does not explicitly describe the specific hardware used to run its experiments, such as CPU or GPU models.
Software Dependencies No The paper states 'We implement our algorithm 1 in MATLAB.' but does not provide a version number for MATLAB or list any other software dependencies with specific versions.
Experiment Setup Yes We choose dimension p = 1000, and let s1 = s2 = 1. The number of samples n varied between 1 and 1000. ... We choose three different matrices Q... We increase n from 1 to 40, and the plot we get is figure 2(b). In this experiment, we choose different dimension p with p = 20, p = 40, p = 80, and p = 160. We let s1 = s2 = 1.