EKSTRASI FITUR DOMAIN WAKTU DAN KLASIFIKASI SINYAL EEG MENGGUNAKAN SVM UNTUK IMAJINASI KATA
Total View This Week0

Institusion

Institut Teknologi Perusahaan Listrik Negara

Author

JANOTTAMA, LAKSAMANA NUGROHO NAGARA
Yosrita, Efy
Yudho, Satrio

Subject

Teknik Informatika 

Datestamp

2022-09-20 03:04:03 

Abstract :
This study aims to classify 8 words, namely ?eat?, ?drink?, ?hungry?, ?thirst?, ?happy?, ?sad?, ?sick?, and ?toilet? in the condition of reading silently and imagining the word. with 0.5 second time domain extraction. In this research case, the author uses a support vector machine model or method with 3 variations of kernel functions, namely the Radial basis function, Linear, and Sigmoid kernels. The results obtained by the author in his research using the support vector machine algorithm with the best kernel, it is found that the results of 8 heart condition words get 22% accuracy with precision (08%, 30%, 35%, 42%, 22%, 0%, 22 %,30%),Recall(12%,50%,20%,12%,47%,0%,24%,22%), F1_score (10%, 37%, 25%, 19%, 30%, 0%, 23%, 25%) while for 8 condition words imagine getting 21% accuracy with precision (25%, 17%, 25%, 26%, 19%, 24%, 07%, 29%), Recall (38 %, 13%, 11%, 30%, 28%, 22%, 25%, 29%), F1_score (30%, 15%, 15%, 28%, 22%, 23%, 11%, 29%) . 

File :
201731232_LaksamanaNugrohoNagaraJanottama_Rev_laksamana nugroho na.pdf