返回
Causal feature selection framework for stable soft sensor modeling based on time-delayed cross mapping
DOI:10.1016/j.aei.2026.104337.png)
摘要
En 中文
● 特征选择考虑了不同时间延迟的因果强度。● 状态空间重建被用来处理相互依赖的变量。● 时间延迟的部分交叉映射被建议用于直接因果推断。● 特征通过因果强度和验证性能自动选择。● 通过直接因果关系选择的特征可以增强模型稳定性。
期刊
IF:
9.9
论文数:
4.4K
被引数:
1.7W
机构
引用论文
An evolutionary deep learning soft sensor model based on random forest feature selection technique for penicillin fermentation process
ISA TRANSACTIONS
IF6.5
Determining embedding dimension for phase-space reconstruction using a geometrical construction使用几何构造确定相空间重建的嵌入维数
A multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor for global prediction of ammonia-nitrogen concentration in wastewater treatment processes
WATER RESEARCH
IF12.4

