Intelligent fault diagnosis of machines with small &amp; imbalanced data: A state-of-the-art review and possible extensions

Zhang, Tianci; Chen, Jinglong; Li, Fudong; Zhang, Kaiyu; Lv, Haixin; He, Shuilong; Xu, Enyong

doi:10.1016/j.isatra.2021.02.042

Cited by 312 publications

(117 citation statements)

References 85 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When constructing weak learners into an ensemble model in many mainstream methods, they will not combine the weak learners directly but get the weighted average of all weak learners. The weight functions are usually relative to the classification results of the base estimators generated in the previous training process [46]. In RUSBoost [37], the pseudo-loss will be calculated to update the weight parameters which will have a great impact on the resampling method in the next iteration and will directly affect the weight of the base estimator trained in this iteration.…”

Section: Contamination Entropy Weight Ensemblementioning

confidence: 99%

ASE: Anomaly Scoring Based Ensemble Learning for Imbalanced Datasets

Liang¹,

Gao²,

Xu³

2022

Preprint

View full text Add to dashboard Cite

Nowadays, many industries have applied classification algorithms to help them solve problems in their business, like finance, medicine, manufacturing industry and so on. However, in real-life scenarios, positive examples only make up a small part of all instances and our datasets suffer from high imbalance ratio which leads to poor performance of existing classification models. To solve this problem, we come up with a bagging ensemble learning framework based on an anomaly detection scoring system. We test out that our ensemble learning model can dramatically improve performance of base estimators (e.g. Decision Tree, Multilayer perceptron, KNN) and is more efficient than other existing methods under a wide range of imbalance ratio, data scale and data dimension.

show abstract

Section: Contamination Entropy Weight Ensemblementioning

confidence: 99%

ASE: Anomaly Scoring Based Ensemble Learning for Imbalanced Datasets

Liang¹,

Gao²,

Xu³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Such issues fall within the scope of zero-shot learning, wherein a model is required to observe and predict samples from a previously unseen class or distribution. [50]. For zero-shot learning to work, there has to be a distinct characteristic of faults and healthy states that is true for previously unseen faults or healthy states, which can be leveraged to assign these distributions to the correct class.…”

Section: Unsupervised Learningmentioning

confidence: 99%

“…Few-shot learning provides an interesting opportunity to learn fault features with only a few instances in a training datasets, as can be the case for many rare faults or components. In the case where no labels exist, supervision algorithms might still be applicable through zero-shot learning [50]. In zero-shot learning, the model seeks to generalise knowledge from seen classes to unseen classes with similar behaviour, much like how humans can see images of house-cats and dogs and then correctly categorize lions to felines and wolves to canines [101,102].…”

Section: Weak Supervisionmentioning

confidence: 99%

Technical Language Supervision for Intelligent Fault Diagnosis in Process Industry

Karl¹,

Taal²,

Schnabel³

et al. 2021

Preprint

View full text Add to dashboard Cite

In the process industry, condition monitoring systems with automated fault diagnosis methods assist human experts and thereby improve maintenance efficiency, process sustainability, and workplace safety. Improving the automated fault diagnosis methods using data and machine learning-based models is a central aspect of intelligent fault diagnosis (IFD). A major challenge in IFD is to develop realistic datasets with accurate labels needed to train and validate models, and to transfer models trained with labeled lab data to heterogeneous process industry environments. However, fault descriptions and work-orders written by domain experts are increasingly digitized in modern condition monitoring systems, for example in the context of rotating equipment monitoring. Thus, domain-specific knowledge about fault characteristics and severities exists as technical language annotations in industrial datasets. Furthermore, recent advances in natural language processing enable weakly supervised model optimization using natural language annotations, most notably in the form of natural language supervision (NLS). This creates a timely opportunity to develop technical language supervision (TLS) solutions for IFD systems grounded in industrial data, for example as a complement to pre-training with lab data to address problems like overfitting and inaccurate out-ofsample generalisation. We surveyed the literature and identify a considerable improvement in the maturity of NLS over the last two years, facilitating applications beyond natural language; a rapid development of weak supervision methods; and transfer learning as a current trend in IFD which can benefit from these developments. Finally, we describe a framework for integration of TLS in IFD which is inspired by recent NLS innovations.

show abstract

“…Targeting the unbalanced data issue in fault diagnosis, researchers have proposed a variety of methods [ 12 – 14 ]. Currently, solutions to unbalanced classification mainly fall into two categories: data-level methods and algorithm-level methods.…”

Section: Introductionmentioning

confidence: 99%

Unbalanced Fault Diagnosis Based on an Invariant Temporal-Spatial Attention Fusion Network

Liu

Yang

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The health status of mechanical bearings concerns the safety of equipment usage. Therefore, it is of crucial importance to monitor mechanical bearings. Currently, deep learning is the mainstream approach for this task. However, in practical situations, the majority of fault samples have the issue of severe class unbalancing, which renders conventional deep learning inapplicable. Targeted at this issue, this paper proposes an invariant temporal-spatial attention fusion network called ITSA-FN for bearing fault diagnosis under unbalanced conditions. First, the proposed method utilizes the invariant temporal-spatial attention representation section, which consists of a pretrained convolutional auto-encoder model, a convolutional block attention module, and a long short-term memory network, to extract independent features and invariant features of spatial-temporal characteristics from input signals. Then, a multilayer perceptron is used to fuse and infer from the extracted features and design a new loss function from the focal loss for network training. Finally, this article validates proposed model’s effectiveness through comparative experiments, ablation studies, and generalization performance experiments.

show abstract

Intelligent fault diagnosis of machines with small & imbalanced data: A state-of-the-art review and possible extensions

Cited by 312 publications

References 85 publications

ASE: Anomaly Scoring Based Ensemble Learning for Imbalanced Datasets

ASE: Anomaly Scoring Based Ensemble Learning for Imbalanced Datasets

Technical Language Supervision for Intelligent Fault Diagnosis in Process Industry

Unbalanced Fault Diagnosis Based on an Invariant Temporal-Spatial Attention Fusion Network

Contact Info

Product

Resources

About