Year: 2025
Type: conference-paper
Abstract: Multivariate Time Series Generation (MTSG) plays a crucial role in time series analysis, supporting tasks such as data augmentation and anomaly detection. While several methods exist for MTSG, recommending the most suitable method for new scenarios remains a significant challenge. Although prior work by [2] provides guidance for selecting MTSG methods, it lacks coverage of recent diffusion-based methods and has limited exploration of channel-independent frameworks. We address these gaps by improving the recommendation guide, highlighting the effectiveness of a central discriminator within the channel-independent framework. Our revised guide makes three key recommendations: 1) VAE-based methods excel on small-scale datasets; 2) a channel-independent framework with the newly designed central discriminator is optimal in most cases; and 3) a diffusion-based method is preferable when ample data and computational resources are available.
Institutions: Xidian University, Guangzhou Institute of Technology, Nio (China)
Language: English
Subfield: Signal Processing
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4410637830