We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Klasifikasi Sinyal EEG pada Sistem BCI Pergerakan Jari Manusia Menggunakan Convolutional Neural Network.
- Authors
Widadi, Rahmat; Widodo, Bongga Arif; Zulherman, Dodi
- Abstract
A Brain-computer interface provides a new communication bridge between the human mind and devices, depending on the accurate classification and identification of EEG signals. The deep learning approach such as Convolutional Neural Network (CNN) for classification has successfully used in many fields, which is potentially used to motor imagery classification. This paper proposes a method that uses CNN for EEG-based motor imagery (MI-EEG) classification. The proposed method consisting of convolution layer and multilayer perceptron was implemented by using TensorFlow 2.0 library (Keras) in Python 3.7. We used MI-EEG 5F from five subjects with a sampling rate of 200 Hz. Testing involves Kfold-cross validation and confusion matrix analysis. Based on the result, kernel size increasing affected accuracy improvement. The best accuracy for the 50 kernels size reaches 51,711%. The proposed method gets better accuracy than the primary reference.
- Subjects
CONVOLUTIONAL neural networks; BRAIN-computer interfaces; CLASSIFICATION; DEEP learning; ELECTROENCEPHALOGRAPHY; PYTHON programming language
- Publication
Techno.com, 2020, Vol 19, Issue 4, p459
- ISSN
1412-2693
- Publication type
Article
- DOI
10.33633/tc.v19i4.4119