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; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]

Non-Convex Projected Gradient Descent for Generalized Low-Rank Tensor Regression

Authors: Han Chen, Garvesh Raskutti, Ming Yuan

JMLR 2019 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We supplement our theoretical results with simulations which show that, under several common settings of generalized low rank tensor regression, the projected gradient descent approach is superior both in terms of statistical error and run-time provided the step-sizes of the projected descent algorithm are suitably chosen.
Researcher Affiliation Academia Han Chen EMAIL Garvesh Raskutti EMAIL Department of Statistics University of Wisconsin Madison Madison, WI 53706, USA Ming Yuan EMAIL Department of Statistics Columbia University New York, NY 10027, USA
Pseudocode Yes Algorithm 1 Projected Gradient Descent 1: Input : data Y, X, parameter space Θ, iterations K, step size η 2: Initialize : k = 0, b T0 Θ 3: for k = 1, 2, . . . , K do 4: gk = b Tk η f( b Tk) (gradient step) 5: b Tk+1 = PΘ(gk) or b Tk+1 = b PΘ(gk) ((approximate) projection step) 7: Output : b TK
Open Source Code No The paper does not provide any explicit statement about releasing source code, a direct link to a code repository, or mention of code in supplementary materials.
Open Datasets No We first describe three different ways of generating random tensor coefficient T with different types of low tensor rank structure... Then we generate covariates {X(i)}n i=1 to be i.i.d random matrices filled with i.i.d N(0, 1) entries. Finally, we simulate three GLM model, the Gaussian linear model, logistic regression and Poisson regression as follows.
Dataset Splits No The paper describes generating synthetic data for simulations but does not specify any train/test/validation splits; rather, new data is generated for each simulation run for comparison of methods.
Hardware Specification No The paper does not provide any specific hardware details like GPU/CPU models, memory amounts, or detailed computer specifications used for running the experiments.
Software Dependencies No The paper mentions using 'a generic cvx solver' in Section 1 and Section 5.3, but does not provide any specific version numbers for this solver or any other software dependencies.
Experiment Setup Yes In all our simualtions, the step-size Ī· is set as a constant specified in each plot. In the first two cases (see cases below), PGD with approximate projection b PΘ3(r ,r ,r ) were applied with different choices of (r , Ī·) while in the third case the PGD with exact projection PΘ2(r ,s ) were adopted with different choices of (r , s , Ī·).