상세정보
학술대회
Locality-based Time Series Data Augmentation for Multi-Sensor Internet Of Things Terminal
- 발행일
- 202210
- 출처
- International Conference on Information and Communication Technology Convergence (ICTC) 2022, pp.1726-1728
- 협약과제
- 21IH4800, 실시간 공공데이터 전달 및 공유 플랫폼 개발(DDS 융합연구단)_본계정, 이정희
- 초록
- Existing approaches for time series data augmentation of deep learning are mostly applied for artificial intelligence applications such as time series classification, time series anomaly detection, and forecasting. In this paper, we propose a locality-based data augmentation for multi-sensor Internet of Things (IoT) terminal. By providing locality-based co-related augmented time series data sets, our approach not only resolves normally collected time series data missing-out problem but also resolves unavoidable error insertion problem of artificial intelligence deep learning compensation.
- KSP 제안 키워드
- Data Augmentation, Data missing, Data sets, Existing Approaches, Internet of thing(IoT), Multi-Sensor, Time Series Classification, Time series data, anomaly detection, artificial intelligence applications, deep learning(DL)