Generalizing across Temporal Domains with Koopman Operators

Authors: Qiuhao Zeng, Wei Wang, Fan Zhou, Gezheng Xu, Ruizhi Pu, Changjian Shui, Christian Gagné, Shichun Yang, Charles X. Ling, Boyu Wang

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

Reproducibility Variable Result LLM Response
Research Type Experimental Through empirical evaluations conducted on synthetic and real-world datasets, we validate the effectiveness of our proposed approach.
Researcher Affiliation Academia 1University of Western Ontario 2Beihang University 3Vector Institute 4Universit e Laval
Pseudocode Yes Algorithm 1: TKNets (one episode)
Open Source Code No The paper does not contain an explicit statement about releasing its source code or a link to a code repository for the described methodology.
Open Datasets Yes We evaluated our algorithm on 6 datasets, 4 of which were collected in real-world scenarios (RMNIST (Ghifary et al. 2015), Portrait (Kumar, Ma, and Liang 2020; Chen and Chao 2021), Cover Type (Kumar, Ma, and Liang 2020)), and FMo W (Christie et al. 2018)).
Dataset Splits No The paper describes how data is divided into domains and how support/query sets are used during training and testing (e.g., 'randomly choose the data sets Si, Si+1 from two consecutive domains as support and query sets'), but it does not provide explicit training, validation, and test split percentages or counts for reproducibility.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used to run the experiments.
Software Dependencies No The paper mentions 'All the baselines and experiments were implemented with Domain Bed package (Gulrajani and Lopez-Paz 2020)', but it does not specify a version number for this package or any other software dependencies.
Experiment Setup No The paper describes the sampling strategy for support and query sets during training ('sample NB samples from each class k in Si') and mentions that experiments were implemented with 'Domain Bed package (Gulrajani and Lopez-Paz 2020) under the same setting', but it does not explicitly list concrete hyperparameter values such as learning rate, batch size, or number of epochs within the main text.