.png){kind=icon}
.png){kind=icon}
Return
Max-Min Robust Principal Component Analysis
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2022.11.092.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Principal Component Analysis (PCA) is a powerful unsupervised dimensionality reduction algorithm, which uses squared '2-norm to cleverly connect reconstruction error and projection variance, and those improved PCA methods only consider one of them, which limits their performance. To alleviate this problem, we propose a novel Max-Min Robust Principal Component Analysis via binary weight, which ingeniously combines reconstruction error and projection variance to learn projection matrix more accurately, and uses '2-norm as evaluation criterion to make the model rotation invariant. In addition, we design binary weight to remove outliers to improve the robustness of model and obtain the ability of anomaly detection. Subsequently, we exploit an efficient iterative optimization algorithm to solve this problem. Extensive experimental results show that our model outperforms related state-of-the-art PCA methods.
Keywords:
Robust dimensionality reduction
Reconstruction
Variance
Anomaly detection
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W