Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

Provable Non-convex Phase Retrieval with Outliers: Median TruncatedWirtinger Flow

Authors: Huishuai Zhang, Yuejie Chi, Yingbin Liang

ICML 2016 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 6. Numerical Experiments In this section, we provide numerical experiments to demonstrate the effectiveness of median-TWF, which corroborates with our theoretical findings.
Researcher Affiliation Academia Huishuai Zhang EMAIL Department of EECS, Syracuse University, Syracuse, NY 13244 USA Yuejie Chi EMAIL Department of ECE, The Ohio State University, Columbus, OH 43210 USA Yingbin Liang EMAIL Department of EECS, Syracuse University, Syracuse, NY 13244 USA
Pseudocode Yes Algorithm 1 Median Truncated Wirtinger Flow (Median TWF)
Open Source Code No The paper does not provide any explicit statement or link indicating the availability of its source code.
Open Datasets No For each pair of (m, n), we generate a signal x N(0, In n), and the measurement vectors ai N(0, In n) i.i.d. for i = 1, . . . , m.
Dataset Splits No The paper generates synthetic data for each trial and does not define training, validation, or specific test splits in the traditional sense of partitioning a pre-existing dataset.
Hardware Specification No The paper does not provide specific details about the hardware specifications (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software details, such as library names with version numbers, needed to replicate the experiment.
Experiment Setup Yes We set the step size in the median-TWF to be a fixed small constant, i.e., µt = 0.2. The rest of the parameters {αy, αh, αl, αu} are set to satisfy ζ1 := max n E h ξ21{|ξ|< 1.01αl or |ξ|> 1.01αl or |ξ|> ζ2 := E ξ21{|ξ|>0.248αh} , (10) 2(ζ1 + ζ2) + p 8/πα 1 h < 1.99 αy 3, where ξ N(0, 1). For example, we set αl = 0.3, αu = 5, αy = 3 and αh = 12, and consequently ζ1 0.24 and ζ2 0.032. ... For both algorithms, a fixed number of iterations T = 500 are run...