Local Hyper-Flow Diffusion

Authors: Kimon Fountoulakis, Pan Li, Shenghao Yang

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate the new technique is significantly better than state-of-the-art methods on both synthetic and real-world data. In Section 6 we perform experiments using both synthetic and real datasets.
Researcher Affiliation Academia Kimon Fountoulakis School of Computer Science University of Waterloo Waterloo, ON, Canada kimon.fountoulakis@uwaterloo.ca Pan Li Department of Computer Science Purdue University West Lafayette, IN, United States panli@purdue.edu Shenghao Yang School of Computer Science University of Waterloo Waterloo, ON, Canada shenghao.yang@uwaterloo.ca
Pseudocode Yes Algorithm 1 Alternating Minimization for (4) Initialization: φ(0) := 0, r(0) := 0, s(0) e := D 1Ae [ d]+ , e E. For k = 0, 1, 2, . . . do: (φ(k+1), r(k+1)) := argmin (φ,r) C e E (φ2 e + 1 σ s(k) e re 2 2) s(k+1) := argmin s e E se r(k+1) e 2 2 e E se d, se,v = 0, v e.
Open Source Code Yes Code that reproduces all results is available at https://github.com/s-h-yang/HFD.
Open Datasets Yes Amazon-reviews ([13,47]). Trivago-clicks ([45]). Florida Bay food network ([8]).
Dataset Splits No The paper does not provide specific details on how datasets were split into training, validation, and test sets. It describes how seed nodes were chosen for evaluation but not the data partitions themselves.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup Yes We set σ = 0.01. For HFD, we initialize the seed mass so that 1 is a constant factor times the volume of the target cluster. We set σ = 0.0001 for HFD and we set the parameters for LH-2.0, LH-1.4 and ACL as suggested by the authors [9].