New OSSE-LSTM framework tackles label-efficient time series classification

Researchers have developed a new framework called Dual-Stream OSSE-LSTM for time series classification, designed to be effective even with a limited number of labeled examples. This method combines a Convolutional Neural Network (CNN) for multi-scale feature extraction with a Bidirectional LSTM for temporal context. To address the need for explainability with scarce data, the system introduces Counterfactual Integrated Gradients (C-IG) for attribution, which helps refine predictions and provides insights into classification decisions. AI

IMPACT This research offers a method to improve time series classification accuracy with limited labeled data, potentially reducing annotation costs in fields like industrial monitoring and healthcare.

RANK_REASON The item is an academic paper detailing a new method for time series classification. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• CORE Recommender
• Counterfactual Integrated Gradients
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• Litmaps
• Omni-Scale CNN
• OSSE-LSTM
• ScienceCast
• scite Smart Citations
• Squeeze-and-Excitation Networks

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →