Application of Recurrent Neural Networks in the Classification Problem of Failures in the Complex Technical Systems Within the Framework of Proactive Maintenance

Demidova, Liliya; Marchev, D.V.

doi:10.21667/1995-4565-2019-69-135-148

Cited by 6 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ELM network uses the Moore-Penrose inversion (Demidova & Marchev, 2019;Huang et al, 2004), due to which the entire architecture of the final neural network consists of one hidden layer. This allows you to increase the learning rate.…”

Section: Extreme Learning Machinementioning

confidence: 99%

Disk Drives Remaining Useful Life Prediction Using the Extreme Learning Machine

Demidova¹

2023

European Proceedings of Computers and Technology

View full text Add to dashboard Cite

The article deals with the problem of predicting the remaining useful life of disk drives using a machine learning model, in particular, using an Extreme Learning Machine (ELM). A method is proposed for improving the values of model quality metrics by generating new features, as well as their selection using a method that implements the calculation of the symmetric Kullback-Leibler Divergence (SKLD). It is shown that a model based on an extreme learning machine and trained on the basis of a dataset formed from the results of generation of new features and their subsequent selection by the SKL method can predict the remaining useful life with an average error of 2.5 days, while model training time is about 6 seconds. The results of a comparative analysis are presented, confirming the efficiency of the proposed model based on ELM. Additionally, the methods for generating features BY and MI are compared, and their shortcomings over SKLD in this case are demonstrated.

show abstract

Section: Extreme Learning Machinementioning

confidence: 99%

Disk Drives Remaining Useful Life Prediction Using the Extreme Learning Machine

Demidova¹

2023

European Proceedings of Computers and Technology

View full text Add to dashboard Cite

show abstract

“…In this case, [36] were used: -Adam's algorithm as an optimization algorithm [37]; -Accuracy indicator as an objective function, where P -number of objects for which the neural network made the correct decision; N -number of objects in the training set.…”

Section: Quality Assessment Of a Trained Neural Networkmentioning

confidence: 99%