Implicit Gaussian process representation of vector fields over arbitrary latent manifolds

Authors: Robert Peach, Matteo Vinao-Carl, Nir Grossman, Michael David, Emma Mallas, David J. Sharp, Paresh A. Malhotra, Pierre Vandergheynst, Adam Gosztolai

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate that RVGP possesses global regularity over the manifold, which allows it to super-resolve and inpaint vector fields while preserving singularities. Furthermore, we use RVGP to reconstruct high-density neural dynamics derived from low-density EEG recordings in healthy individuals and Alzheimer s patients. We show that vector field singularities are important disease markers and that their reconstruction leads to a classification accuracy of disease states comparable to high-density recordings.
Researcher Affiliation Academia Robert L. Peach* University Hospital W urzburg peach r@ukw.de Matteo Vinao-Carl*, Nir Grossman, Michael David Imperial College London (* Indicates equal contribution) Emma Mallas, David Sharp, Paresh A. Malhotra Imperial College London Pierre Vandergheynst, Adam Gosztolai EPFL adam.gosztolai@epfl.ch
Pseudocode No The paper describes the proposed method mathematically and textually but does not include any explicitly labeled pseudocode or algorithm blocks.
Open Source Code No The paper states, "...our method is readily compatible with inducing point methods, which we provide an implementation of." (Section 4.2). However, it does not provide a specific link or explicit statement about the code being publicly released.
Open Datasets No The paper mentions collecting EEG data from Hammersmith Hospital in London and the UKDRI Care Research & Technology Centre (Section 5.2, Appendix A.1, A.7). However, it does not provide concrete access information (link, DOI, repository, or citation to a public dataset) for this data.
Dataset Splits Yes For binary classification, we used a linear support vector machine and computed the accuracy and ROC curves for each approach separately using 10-fold cross-validation.
Hardware Specification No The paper does not mention any specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments.
Software Dependencies No The paper mentions "MATLAB R2023a" and "eeglab (Delorme & Makeig, 2004)". While MATLAB has a version, EEGLAB is a toolbox cited by its publication, not given a specific software version. No other specific software dependencies with versions are listed.
Experiment Setup No The paper describes the general setup of experiments, such as using K-nearest neighbour graph (K=5) for kernel construction, but it does not specify concrete hyperparameters like learning rates, batch sizes, number of epochs, or optimizer settings for RVGP training.