Towards Robust Deterministic and Probabilistic Modeling for Predictive Learning

Authors: Xuesong Nie, Haoyuan Jin, Vijayakumar Bhagavatula, Xiaofeng Liu

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments demonstrate the superiority of DDP across diverse scenario evaluations. ... Extensive experiments show that DDP achieves state-of-the-art performance across various real-world scenes. ... We demonstrate the effectiveness of the DDP model with multi-scenario evaluations. ... In this section, we further perform extensive ablation studies to study the components effectiveness in our DDP.
Researcher Affiliation Academia Xuesong Nie1,2 , Haoyuan Jin1 , Vijayakumar Bhagavatula3 and Xiaofeng Liu2, 1Zhejiang University 2Yale University 3Carnegie Mellon University EMAIL, EMAIL, EMAIL
Pseudocode No The paper describes methods and algorithms using mathematical equations and textual explanations, but it does not contain a clearly labeled "Pseudocode" or "Algorithm" block, nor does it present structured steps in a code-like format.
Open Source Code No The paper does not contain any explicit statement about releasing source code or provide any links to a code repository.
Open Datasets Yes 4.1 Human Motions: UCF Sports Dataset and Setup. UCF Sports [Rodriguez et al., 2008] ... 4.2 Synthetic Motions: Moving MNIST Dataset and Setup. The Moving MNIST [Srivastava et al., 2015] dataset is constructed ... 4.3 Driving Scenes: KITTI&Caltech Dataset and Setup. The KITTI&Caltech [Geiger et al., 2013; Doll ar et al., 2009] dataset ... 4.4 Traffic Flow: Taxi BJ Dataset and Setup. Taxi BJ [Zhang and others, 2017] comprises taxi GPS trajectory data ... 4.5 Global Climate: Weather Bench Dataset and Setup. Weather Bench [Rasp et al., 2020] contains climatic data ...
Dataset Splits Yes UCF Sports ... using 6,288 sequences for training and 752 for testing. ... Moving MNIST ... There are 10,000 sequences for training and 10,000 for testing. ... KITTI&Caltech ... we train the model on the KITTI [Geiger et al., 2013] dataset and evaluate it against the Caltech Pedestrian [Doll ar et al., 2009] dataset. ... Weather Bench ... using 2010-2015 for training, 2016 for validation, and 2017-2018 for testing.
Hardware Specification Yes Our method uses Py Torch on an NVIDIA A100 GPU
Software Dependencies No Our method uses Py Torch on an NVIDIA A100 GPU, training with 16-sequence minibatches, the Adam optimizer, and the One Cycle scheduler. While PyTorch is mentioned, a specific version number is not provided, nor are versions for other software components.
Experiment Setup Yes Implementation Details. Our method uses Py Torch on an NVIDIA A100 GPU, training with 16-sequence minibatches, the Adam optimizer, and the One Cycle scheduler. We apply a weight decay of 5e 2 and select learning rates from {1e 2, 5e 3, 1e 3} for stability. We use the MSE loss to supervise training and stochastic depth for regularization.