High-dimensional Location Estimation via Norm Concentration for Subgamma Vectors

Authors: Shivam Gupta, Jasper C.H. Lee, Eric Price

ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We perform experimental validation on a synthetic high-dimensional example. and Figure 3. Error scaled by N for different algorithms, for a synthetic Gaussian mixture.
Researcher Affiliation Academia 1The University of Texas at Austin 2Department of Computer Sciences and Institute for Foundations of Data Science, University of Wisconsin-Madison.
Pseudocode Yes Algorithm 1 Local smoothed MLE for one dimension, Algorithm 2 Global smoothed MLE for one dimension, Algorithm 3 High-dimensional Local MLE, Algorithm 4 High-dimensional Global MLE
Open Source Code Yes Our implementation is available here: https://github.com/shivamgupta2/High-dimensional-location
Open Datasets No The paper uses a synthetic dataset described as 'a mixture of three gaussians' which is generated for the experiments, with no concrete access information provided.
Dataset Splits No The paper does not provide specific dataset split information (percentages, sample counts, or citations to predefined splits).
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory amounts) used for running its experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers (e.g., library names with versions).
Experiment Setup Yes We consider three algorithms: our algorithm with smoothing radius 0.1;... For our algorithm with R = 0.01I... Newton s method (i.e., our algorithm except with R = 0 and multiple steps; we use 10 steps).