CARE: Modeling Interacting Dynamics Under Temporal Environmental Variation

Authors: Xiao Luo, Haixin Wang, Zijie Huang, Huiyu Jiang, Abhijeet Gangan, Song Jiang, Yizhou Sun

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Comprehensive experiments on four datasets demonstrate the effectiveness of our proposed CARE compared with several state-of-the-art approaches.
Researcher Affiliation Academia 1University of California, Los Angeles, 2Peking University, 3University of California, Santa Barbara
Pseudocode Yes E Algorithm The whole learning algorithm of CARE is summarized in Algorithm 1.
Open Source Code No The paper does not provide an explicit statement about the release of source code for the described methodology, nor does it include any links to a code repository.
Open Datasets Yes We evaluate our proposed CARE on two particle-based simulation datasets with temporal environmental variations, i.e., Lennard-Jones Potential [47] and 3-body Stillinger-Weber Potential [4]. ... We employ two popular mesh-based simulation datasets, i.e., Cylinder Flow, and Airfoil. Cylinder Flow ... by Open Foam [22]. Airfoil is generated in a similar manner ... by Open Foam [22].
Dataset Splits Yes To ensure the accuracy of our results, we use a rigorous data split strategy, where first 80% of the samples are reserved for training purposes and the remaining 10% are set aside for testing and validating, separately.
Hardware Specification Yes All experiments are conducted on a single NVIDIA A100 GPU.
Software Dependencies No We employ the fourth-order Runge-Kutta method as in the torchdiffeq Python package [24], using Py Torch [40]. While software is mentioned, specific version numbers for PyTorch and torchdiffeq are not provided.
Experiment Setup Yes We set the latent dimension to 256 and the dropout rate to 0.2. For optimization, we use the Adam optimizer with weight decay by mini-batch stochastic gradient descent, setting the learning rate to 0.01.