SelF-Rocket achieves state-of-the-art accuracy in time series classification

Researchers have developed SelF-Rocket, a novel method for time series classification that builds upon the MiniRocket framework. This new approach dynamically selects optimal input representations and pooling operators during training. SelF-Rocket has demonstrated state-of-the-art performance on the University of California Riverside (UCR) benchmark datasets for time series classification. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel approach to time series classification, potentially improving performance on benchmark datasets.

RANK_REASON This is a research paper presenting a new method for time series classification.

Read on arXiv cs.LG →

• SelF-Rocket
• MiniRocket
• University of California Riverside
• UCR

• paper
• other