KLASIFIKASI PARASIT MALARIA PLASMODIUM VIVAX MENGGUNAKAN JARINGAN SYARAF TIRUAN BACKPROPAGATION
| Main Author: | Septinia, Indriani |
|---|---|
| Format: | Thesis NonPeerReviewed Book |
| Bahasa: | ind |
| Terbitan: |
, 2015
|
| Subjects: | |
| Online Access: |
https://eprints.untirta.ac.id/9378/1/KLASIFIKASI%20PARASIT%20MALARIA%20PLASMODIUM%20VIVAX.pdf https://eprints.untirta.ac.id/9378/ https://ft.untirta.ac.id |
Daftar Isi:
- World Health Organization (WHO) states that malaria is one of serious global disease, malaria attack a half of the world population. Some of developing country have a difficult problem to identifying malaria parasite because an expensive cost and lack of diagnostic expert. Therefore it’s required an efficient alternative method of diagnosis. This research aims to classify the image of vivax plasmodium parasite that contain of malaria parasite. 90 of the image is clssified according to the phases of the plasmodium development, there are 89 image of training and 1 image on testing, the phase of malaria named Trophozoite, Schizonts, and Gametocyte. The feature extraction that used as the input is Mean, Standard Deviation, Kurtosis, Skewness dan Entropy from the color histogram, grayscale histogram, and level saturation histogram. The parameters of classification using 2 of hidden layer and 1 output layer. The first hidden layer using 8 of neurons, second hidden layer using 5 of neurons, and 3 neurons for output layer. Then the process of classification vivax plasmodium parasite was carried out using the Backpropagation from Neural Network. The results of accuracy from each phase is 86.7% for Trophozoite, 83.3% for Schizonts, and 80% for Gametocyte, with a total of 83.3% accuracy. A sensitivity of trophozoite stage is 0.7, schizonts 0.714, and gametocyte 0,727. Specificity of the trophozoite stage is 0.075, schizonts 0.09, and gametocyte 0.1, with 83.3% of effectiveness system. Keyword: vivax malaria, neural network, feature extraction, mean, standard deviasi, kurtosis, skewness, entropy, histogram.