Skip to main content
Skip to footer
Contact Contact

Automated label generation for time series classification with representation learning: Reduction of label cost for training

You have these already downloaded

We have sent you a copy of the report to your email again.

Abstract

Time series generated by end users, edge devices, and different wearables are mostly unlabelled. We propose a method to auto-generate labels of unlabelled time series, exploiting very few representative labelled time series. Our method is based on representation learning using auto-encoded compact sequence, with a choice of best distance measure. It performs self-correction in iterations by learning latent structure as well as synthetically boosting representative time series using variational auto-encoder to improve the quality of labels. 

We have experimented with UCR and UCI archives, public real-world univariate and multivariate time series taken from different application domains. Experimental results demonstrate that the proposed method is very close to the performance achieved by fully supervised classification. It not only produces close-to-benchmark results but outperforms the benchmark performance in some cases.

Research area: Embedded devices and intelligent systems

Authors: Soma Bandyopadhyay, Anish Datta, and Arpan Pal

Conference/event: International Joint Conference on Artificial Intelligence

Conference date: July, 2021

Read More

Related reading

TCS Research

Digital Twin of the Human Skin for In silico Testing

Article | 20 Sep 2023
TCS Research

INAE honors two eminent TCS scientists 

Article | 06 Sep 2023
TCS Research Dheeraj Chahal Dr. Mayank Mishra

AI workload migration to the cloud 

White Paper | 30 Aug 2023
TCS Research Dr. Jayavardhana Gubbi Dr Balamuralidhar Purushothaman Arpan Pal Mahesh Rangarajan

AI-powered secure HyperOps system for emerging business situations

White Paper | 30 Aug 2023

Transformation starts here

Ready to step into the future of retail?

Talk to our experts

Find out more