# Self-learning sparse PCA for multimode process monitoring

arXiv: https://arxiv.org/abs/2108.03449 · HTML: https://mlsanity.com/paper/2108.03449/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 34 references that are in the map.
- First posted on arXiv 7 Aug 2021.

## Builds on (most cited first, 11 of 11)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Continual Lifelong Learning with Neural Networks: A Review](https://mlsanity.com/paper/1802.07569/) (2018)
- [Lifelong Learning with Dynamically Expandable Networks](https://mlsanity.com/paper/1708.01547/) (2017)
- [Task-Free Continual Learning](https://mlsanity.com/paper/1812.03596/) (2018)
- [Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization](https://mlsanity.com/paper/1802.01569/) (2018)
- [The proximal augmented Lagrangian method for nonsmooth composite optimization](https://mlsanity.com/paper/1610.04514/) (2016)
- [An improved mixture of probabilistic PCA for nonlinear data-driven process monitoring](https://mlsanity.com/paper/2012.06830/) (2020)
- [Monitoring multimode processes: a modified PCA algorithm with continual learning ability](https://mlsanity.com/paper/2012.07044/) (2020)
- [Process monitoring based on orthogonal locality preserving projection with maximum likelihood estimation](https://mlsanity.com/paper/2012.07021/) (2020)
- [Monitoring nonstationary processes based on recursive cointegration analysis and elastic weight consolidation](https://mlsanity.com/paper/2101.08579/) (2021)

## Abstract (arXiv)

This paper proposes a novel sparse principal component analysis algorithm with self-learning ability for successive modes, where synaptic intelligence is employed to measure the importance of variables and a regularization term is added to preserve the learned knowledge of previous modes. Different from traditional multimode monitoring methods, the monitoring model is updated based on the current model and new data when a new mode arrives, thus delivering prominent performance for sequential modes. Besides, the computation and storage resources are saved in the long run, because it is not necessary to retrain the model from scratch frequently and store data from previous modes. More importantly, the model furnishes excellent interpretability owing to the sparsity of parameters. Finally, a numerical case and a practical pulverizing system are adopted to illustrate the effectiveness of the proposed algorithm.