Identifikasi Jenis Kanker Darah (Leukemia) Terhadap Pengaruh Parameter Kernel Ssupport Vector Machine Dan Ekstraksi Ciri Dengan Rantai Markov Orde 2

Al Faroby, Moh. Hamim Zajuli (2018) Identifikasi Jenis Kanker Darah (Leukemia) Terhadap Pengaruh Parameter Kernel Ssupport Vector Machine Dan Ekstraksi Ciri Dengan Rantai Markov Orde 2. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
06111340000030_Undergraduate_Theses.pdf - Published Version
Restricted to Repository staff only

Download (3MB) | Request a copy

Abstract

Penelitian terhadap sekuens DNA/RNA leukemia dapat mempermudah dan mempercepat bagi tenaga medis dalam mendiagnosa jenis leukemia. Data leukemia diolah dengan mengekstrasi ciri data sekuens berupa tipe string ke bentuk numerik dengan menggunakan metode rantai Markov orde 2. Hasil dari ekstraksi ciri yaitu 64 ciri dari DNA/RNA leukemia dalam bentuk matriks Markovian. Machine Learning digunakan dalam proses klasifikasi dengan pendekatan metode Support Vector Machine (SVM). Pada saat mengklasifikasikan, penelitian ini menggunakan 40 data training sebagai pengklasifikasi data dan 25 data tes. SVM membagi kelas dari DNA/RNA kanker darah menjadi empat bagian yaitu Acute Myeloid Leukemia, Acute Lymphocytic Leukemia, Chronic Myelogenous Leukemia dan Chronic Lymphocytic Leukemia. Parameter-parameter pada SVM dilakukan uji coba untuk mendapaktan nilai optimal dimana SVM dapat bekerja dengan baik pada data leukemia. Hasil dari penelitian berupa gambar data latih pada SVM. Gambar berupa grafik 2D dimana 64 ciri dari DNA/RNA leukemia akan dikompres dengan menggunakan Principle Component Analysis (PCA). Hasil penelitian yang lain berupa prediksi yang divalidasi dengan percobaan terhadap data test. Serta dibandingkan nya hasil dari tiga kernel yaitu linear, gaussian dan polynomial. Percobaan tersebut digunakan untuk mengetahui performa kernel mana yang paling bagus digunakan pada data sekuens DNA/RNA leukemia. ============== Research on DNA / RNA leukemia sequences can facilitate and accelerate for medical personnel in diagnosing leukemia types. Leukemia data were extracted by extracting sequence data characteristic of string type to numeric form by using Markov 2 chain method 2. The result of characteristic extraction is 64 characteristic of DNA / RNA leukemia in the form of Markovian matrix. Machine Learning is used in the classification process with the approach method of Support Vector Machine (SVM). At the time of classification, this research uses 40 training data as data classifier and 25 test data. SVM divides the class from DNA / RNA blood cancer into four parts namely Acute Myeloid Leukemia, Acute Lymphocytic Leukemia, Chronic Myelogenous Leukemia and Chronic Lymphocytic Leukemia. The parameters in the SVM were tested for obtaining optimal values in which SVM worked well on leukemia data. The result of the research is the image of trainer data on SVM. The image is a 2D graph where 64 features of DNA / RNA leukemia will be compressed using Principle Component Analysis (PCA). Other research results are validated predictions with experiments on the test data. And compared to the results of three kernels are linear, gaussian and polynomial. The experiment was used to determine which kernel performance is best used in DNA / RNA leukemia sequence data.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Rantai Markov; Machine Learning; Support Vector Machine; Kanker Darah; Markov Chain; Machine Learning; Support Vector Machine; Blood Cancer.
Subjects: Q Science > QA Mathematics > QA274.7 Markov processes--Mathematical models.
Q Science > QA Mathematics > QA278.5 Principal components analysis. Factor analysis.
Q Science > QA Mathematics > QA75 Electronic computers. Computer science. EDP
Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Q Science > QR Microbiology > QR 201.T84 Tumors. Cancer
Divisions: Faculty of Mathematics and Science > Mathematics > (S1) Undergraduate Theses
Depositing User: Alfarobya Moh Hamim Zajuli
Date Deposited: 26 Mar 2018 03:36
Last Modified: 26 Mar 2018 03:36
URI: http://repository.its.ac.id/id/eprint/50701

Actions (login required)

View Item View Item