ComS2T: A complementary spatiotemporal learning system for data-adaptive
model evolution
- URL: http://arxiv.org/abs/2403.01738v1
- Date: Mon, 4 Mar 2024 05:31:29 GMT
- Title: ComS2T: A complementary spatiotemporal learning system for data-adaptive
model evolution
- Authors: Zhengyang Zhou, Qihe Huang, Binwu Wang, Jianpeng Hou, Kuo Yang, Yuxuan
Liang, Yang Wang
- Abstract summary: We introduce a prompt-based complementary learning termed ComS2T, to empower the evolution of models for data adaptation.
We disentangle first two disjoint structures into stable and dynamic weights, and then train spatial and temporal prompts by characterizing distribution of main observations.
This data-adaptive prompt mechanism, combined with a two-stage training process, facilitates fine-tuning of the neural architecture conditioned on prompts.
- Score: 20.525608301451687
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Spatiotemporal (ST) learning has become a crucial technique to enable smart
cities and sustainable urban development. Current ST learning models capture
the heterogeneity via various spatial convolution and temporal evolution
blocks. However, rapid urbanization leads to fluctuating distributions in urban
data and city structures over short periods, resulting in existing methods
suffering generalization and data adaptation issues. Despite efforts, existing
methods fail to deal with newly arrived observations and those methods with
generalization capacity are limited in repeated training. Motivated by
complementary learning in neuroscience, we introduce a prompt-based
complementary spatiotemporal learning termed ComS2T, to empower the evolution
of models for data adaptation. ComS2T partitions the neural architecture into a
stable neocortex for consolidating historical memory and a dynamic hippocampus
for new knowledge update. We first disentangle two disjoint structures into
stable and dynamic weights, and then train spatial and temporal prompts by
characterizing distribution of main observations to enable prompts adaptive to
new data. This data-adaptive prompt mechanism, combined with a two-stage
training process, facilitates fine-tuning of the neural architecture
conditioned on prompts, thereby enabling efficient adaptation during testing.
Extensive experiments validate the efficacy of ComS2T in adapting to various
spatiotemporal out-of-distribution scenarios while maintaining efficient
inference capabilities.
Related papers
- SMA-Hyper: Spatiotemporal Multi-View Fusion Hypergraph Learning for Traffic Accident Prediction [2.807532512532818]
Current data-driven models often struggle with data sparsity and the integration of diverse urban data sources.
We introduce a deep dynamic learning framework designed for traffic accident prediction.
It incorporates dual adaptive graph learning mechanisms that enable high-order cross-regional learning.
It also employs an advance attention mechanism to fuse multiple views of accident data and urban functional features.
arXiv Detail & Related papers (2024-07-24T21:10:34Z) - Recurrent Deep Kernel Learning of Dynamical Systems [0.5825410941577593]
Digital twins require computationally-efficient reduced-order models (ROMs) that can accurately describe complex dynamics of physical assets.
We propose a data-driven, non-intrusive deep kernel learning (SVDKL) method to discover low-dimensional latent spaces from data.
Results show that our framework is capable of (i) denoising and reconstructing measurements, (ii) learning compact representations of system states, (iii) predicting system evolution in low-dimensional latent spaces, and (iv) modeling uncertainties.
arXiv Detail & Related papers (2024-05-30T07:49:02Z) - Multi-Modality Spatio-Temporal Forecasting via Self-Supervised Learning [11.19088022423885]
We propose a novel MoST learning framework via Self-Supervised Learning, namely MoSSL.
Results on two real-world MoST datasets verify the superiority of our approach compared with the state-of-the-art baselines.
arXiv Detail & Related papers (2024-05-06T08:24:06Z) - Hybrid Transformer and Spatial-Temporal Self-Supervised Learning for
Long-term Traffic Prediction [1.8531577178922987]
We propose a model that combines hybrid Transformer and self-supervised learning.
The model enhances its adaptive data augmentation by applying data augmentation techniques at the sequence-level of the traffic.
We design two self-supervised learning tasks to model the temporal and spatial dependencies, thereby improving the accuracy and ability of the model.
arXiv Detail & Related papers (2024-01-29T06:17:23Z) - TWINS: A Fine-Tuning Framework for Improved Transferability of
Adversarial Robustness and Generalization [89.54947228958494]
This paper focuses on the fine-tuning of an adversarially pre-trained model in various classification tasks.
We propose a novel statistics-based approach, Two-WIng NormliSation (TWINS) fine-tuning framework.
TWINS is shown to be effective on a wide range of image classification datasets in terms of both generalization and robustness.
arXiv Detail & Related papers (2023-03-20T14:12:55Z) - Gait Recognition in the Wild with Multi-hop Temporal Switch [81.35245014397759]
gait recognition in the wild is a more practical problem that has attracted the attention of the community of multimedia and computer vision.
This paper presents a novel multi-hop temporal switch method to achieve effective temporal modeling of gait patterns in real-world scenes.
arXiv Detail & Related papers (2022-09-01T10:46:09Z) - Spatio-Temporal Graph Few-Shot Learning with Cross-City Knowledge
Transfer [58.6106391721944]
Cross-city knowledge has shown its promise, where the model learned from data-sufficient cities is leveraged to benefit the learning process of data-scarce cities.
We propose a model-agnostic few-shot learning framework for S-temporal graph called ST-GFSL.
We conduct comprehensive experiments on four traffic speed prediction benchmarks and the results demonstrate the effectiveness of ST-GFSL compared with state-of-the-art methods.
arXiv Detail & Related papers (2022-05-27T12:46:52Z) - Averaging Spatio-temporal Signals using Optimal Transport and Soft
Alignments [110.79706180350507]
We show that our proposed loss can be used to define temporal-temporal baryechecenters as Fr'teche means duality.
Experiments on handwritten letters and brain imaging data confirm our theoretical findings.
arXiv Detail & Related papers (2022-03-11T09:46:22Z) - Leveraging the structure of dynamical systems for data-driven modeling [111.45324708884813]
We consider the impact of the training set and its structure on the quality of the long-term prediction.
We show how an informed design of the training set, based on invariants of the system and the structure of the underlying attractor, significantly improves the resulting models.
arXiv Detail & Related papers (2021-12-15T20:09:20Z) - Learning to Continuously Optimize Wireless Resource in a Dynamic
Environment: A Bilevel Optimization Perspective [52.497514255040514]
This work develops a new approach that enables data-driven methods to continuously learn and optimize resource allocation strategies in a dynamic environment.
We propose to build the notion of continual learning into wireless system design, so that the learning model can incrementally adapt to the new episodes.
Our design is based on a novel bilevel optimization formulation which ensures certain fairness" across different data samples.
arXiv Detail & Related papers (2021-05-03T07:23:39Z) - Neural Spatio-Temporal Point Processes [31.474420819149724]
We propose a new class of parameterizations for point-trivial processes which leverage Neural ODEs as a computational method.
We validate our models on data sets from a wide variety of contexts such as seismology, epidemiology, urban mobility, and neuroscience.
arXiv Detail & Related papers (2020-11-09T17:28:23Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.