Unlinked Monotone Regression

Authors: Fadoua Balabdaoui, Charles R. Doss, Cécile Durot

JMLR 2021 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We develop a second order algorithm for its computation, and we demonstrate its use on synthetic data. Finally, we apply our method (in a fully data driven way, without knowledge of the error distribution) on longitudinal data from the US Consumer Expenditure Survey.
Researcher Affiliation Academia Fadoua Balabdaoui EMAIL Seminar f ur Statistik ETH, Zurich R amistrasse 101 8092, Zurich, Switzerland, Charles R. Doss EMAIL School of Statistics University of Minnesota Minneapolis, MN 55455, USA, C ecile Durot EMAIL Modal x Universit e Paris Nanterre F92000, Nanterre, France
Pseudocode Yes Algorithm 1: Active set algorithm, Algorithm 2: Activate constraints: group (approximately) non-unique entries in m
Open Source Code No The paper does not contain an explicit statement about the release of source code for the methodology described, nor does it provide a link to a code repository.
Open Datasets Yes Finally, we apply our method (in a fully data driven way, without knowledge of the error distribution) on longitudinal data from the US Consumer Expenditure Survey. Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek. IPUMS, USA: Version 10.0 [dataset]. Minneapolis, MN: IPUMS, 2020.
Dataset Splits No For synthetic data, the paper mentions sample sizes n=100 and n=1000 but does not specify train/test/validation splits. For the CEX data, it describes a selection criterion for individuals (surveyed in both quarter 2 and quarter 3) but no explicit data splits for experimental reproduction.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running its experiments.
Software Dependencies No The paper discusses algorithms and cites references for optimization methods, but does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment.
Experiment Setup Yes Our simulations were performed taking both Laplace and Gaussian errors with standard deviation 1 (and both unlinked methods are well-specified). The error distribution is assumed to be Laplace distributed with λ = p bσ2/2 where bσ2 is the empirical variance of ϵ1, . . . , ϵ2164. We used a four component Gaussian mixture (with no variance constraint) to approximate the error distribution... The estimate has weights (.56, .05, .33, .06), means/locations ( 345, 416, 326, 1710) and standard deviations (322, 75, 562, 1286).