Ascertainment-adjusted parameter estimation approach to improve robustness against misspecification of health monitoring methods
详细信息 查看全文
- 作者：P. Juesasa ; E. Ramassoa ; b ; emmanuel.ramasso@femto-st.fr" class="auth_mail" title="E-mail the corresponding author
- 关键词：Time-series ; Noisy and uncertain labels ; Weighted distribution theory ; Health monitoring ; CMAPSS datasets
- 刊名：Mechanical Systems and Signal Processing
- 出版年：2016
- 出版时间：15 December 2016
- 年：2016
- 卷：81
- 期：Complete
- 页码：387-401
- 全文大小：1869 K
文摘
- •
Parameters learning based on corrupted and noisy labels.
- •
Method applicable to models with latent variables.
- •
Practical demonstrations on a simulated datasets with different models.
- •
Validation on 700 degradation trajectories from CMAPSS datasets.
- •
Comparison with other models and possible relation with Bayesian learning.