Researchers have introduced msData, a new dataset designed to improve time series foundation models (TSFMs) by incorporating millisecond-resolution data from an operational 5G wireless network. Current TSFM datasets predominantly use lower-frequency data, limiting their effectiveness with high-frequency signals. msData aims to bridge this gap, offering data with sampling intervals as fine as milliseconds and introducing the wireless network domain to complement existing energy and finance datasets. Initial benchmarking indicates that most TSFM configurations perform poorly on this high-frequency data, highlighting the need for such datasets to enhance model pre-training, fine-tuning, and generalization capabilities. AI
IMPACT This dataset could enable TSFMs to better handle high-frequency data, improving their performance in real-world applications like network traffic forecasting.
RANK_REASON The item is a research paper introducing a new dataset for AI models. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
