2019
DOI: 10.1109/tii.2019.2925624
|View full text |Cite
|
Sign up to set email alerts
|

Soft Computing-Based EEG Classification by Optimal Feature Selection and Neural Networks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
5

Citation Types

0
50
0
1

Year Published

2019
2019
2022
2022

Publication Types

Select...
7
3

Relationship

1
9

Authors

Journals

citations
Cited by 91 publications
(55 citation statements)
references
References 14 publications
0
50
0
1
Order By: Relevance
“…Therefore, one idea is to use the artificial features to optimize the feature extraction capability of neural networks, thus retaining the advantages of neural networks and reducing noise emergency interference. For example, some DE-based methods have achieved state of art results in emotion recognition tasks and PSD-based methods are widely used in signal processing areas, which make these two features representative [19], [20]. These results illustrate the effectiveness and potential of artificial features.…”
Section: Introductionmentioning
confidence: 94%
“…Therefore, one idea is to use the artificial features to optimize the feature extraction capability of neural networks, thus retaining the advantages of neural networks and reducing noise emergency interference. For example, some DE-based methods have achieved state of art results in emotion recognition tasks and PSD-based methods are widely used in signal processing areas, which make these two features representative [19], [20]. These results illustrate the effectiveness and potential of artificial features.…”
Section: Introductionmentioning
confidence: 94%
“…After noise removal, features are extracted, and suitable features are selected that give high interclass variance and low intraclass variance [23]. Researchers [5]- [7], [10]- [14], [24] have extracted handcrafted features in both temporal and spectral features for predicting epileptic seizures. Temporal features include the first four statistical moments [25], [26], entropy [27], approximate entropy [25], Hjorth parameters [28] and Lyapunov exponents [29].…”
Section: Introductionmentioning
confidence: 99%
“…Many investigations have developed models implementing FS to identify the cognitive workload using the physiological signal’s information. In [ 12 ], it is shown that soft computing-based EEG classification by extracting and then selecting optimal features can produce better results. The system displays an accuracy of 93.05% and 85.00%, obtaining a low performance in real-time environments.…”
Section: Introductionmentioning
confidence: 99%