Online and stochastic optimization beyond Lipschitz continuity: A Riemannian approach

Authors: Kimon Antonakopoulos, E. Veronica Belmega, Panayotis Mertikopoulos

ICLR 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we supplement our theoretical analysis with numerical experiments in Poisson inverse problems.
Researcher Affiliation Academia Kimon Antonakopoulos Inria, Univ. Grenoble Alpes, CNRS, Grenoble INP, LIG 38000 Grenoble, France kimon.antonakopoulos@inria.fr E. Veronica Belmega ETIS UMR8051, CY University, ENSEA, CNRS, F-95000, Cergy, France belmega@ensea.fr Panayotis Mertikopoulos Inria, Univ. Grenoble Alpes, CNRS, Grenoble INP, LIG 38000 Grenoble, France panayotis.mertikopoulos@imag.fr
Pseudocode No The paper describes algorithms (FTRL and OMD) using mathematical formulations, but it does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets No The paper mentions using a '384 × 384 test image contaminated with Poisson noise' for a 'Poisson denoising problem', implying a specific image (Lena test image is mentioned in the caption of Fig 1). However, it does not provide concrete access information (link, DOI, repository, or formal citation with authors/year) for a publicly available or open dataset.
Dataset Splits No The paper mentions 'minibatch size (n = 256)' and a 'test image' but does not provide specific dataset split information (exact percentages, sample counts, or detailed splitting methodology) for training, validation, or testing.
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 provide specific ancillary software details (e.g., library or solver names with version numbers).
Experiment Setup Yes All algorithms were run with stochastic gradients drawn with the same minibatch size (n = 256) and a step-size of the form γt ∝ 1/√t.