SaSDim:Self-Adaptive Noise Scaling Diffusion Model for Spatial Time Series Imputation

Authors: Shunyang Zhang, Senzhang Wang, Xianzhen Tan, Renzhi Wang, Ruochen Liu, Jian Zhang, Jianxin Wang

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments conducted on three real-world spatial time series datasets verify the effectiveness of Sa SDim.
Researcher Affiliation Academia Central South University {224712166, szwang, 224712155, rzwang, 234712122, jianzhang, jxwang}@csu.edu.cn,
Pseudocode Yes Algorithm 1 Training of Sa SDi M and Algorithm 2 Imputation (Sampling) with Sa SDi M
Open Source Code No The paper does not contain any explicit statement or link indicating the public release of source code for the described methodology.
Open Datasets Yes We evaluate the performance of our model on three spatial time series datasets, METR-LA, AQI-36, and PEMS-BAY. AQI-36 is collected from 36 AQI sensors distributed across the city of Beijing. This dataset serves as a widely recognized benchmark for imputation techniques and includes a mask used for evaluation that simulates the distribution of actual missing data [Yi et al., 2016].
Dataset Splits Yes For the two datasets METR-LA and PEMS-BAY, we partition the entire data into training, validation, and testing sets by a ratio of 8 : 1: 1.
Hardware Specification No This work was carried out in part using computing resources at the High-Performance Computing Center of Central South University.
Software Dependencies No The model is implemented using Pytorch and trained in an end-to-end manner using Adam with a learning rate of 0.001.
Experiment Setup Yes Table 2: The settings of Sa SDi M for the three datasets. Description AQI-36 METR-LA PEMS-BAY Batch size 16 16 16 Time length L 24 24 24 Epochs 200 200 200 Learning rate 0.001 0.001 0.001 Channel size d 64 64 64 Minimum noise level β1 0.0001 0.0001 0.0001 Maximum noise level βT 0.5 0.2 0.2 Diffusion steps T 50 50 50